From a983957b217aa217edf821785b7727096e04048d Mon Sep 17 00:00:00 2001 From: Gabriele Gristina Date: Tue, 9 May 2023 18:52:20 +0200 Subject: [PATCH 1/3] Dependencies: Updated LZMA SDK to 22.01 --- deps/LZMA-SDK/C/7zTypes.h | 20 +- deps/LZMA-SDK/C/7zVersion.h | 10 +- deps/LZMA-SDK/C/7zip_gcc_c.mak | 62 +- deps/LZMA-SDK/C/Aes.c | 10 +- deps/LZMA-SDK/C/Alloc.c | 10 +- deps/LZMA-SDK/C/Alloc.h | 9 +- deps/LZMA-SDK/C/CpuArch.c | 68 +- deps/LZMA-SDK/C/CpuArch.h | 44 +- deps/LZMA-SDK/C/DllSecur.c | 16 +- deps/LZMA-SDK/C/LzFind.c | 1007 +++++++++++++++------- deps/LZMA-SDK/C/LzFind.h | 41 +- deps/LZMA-SDK/C/LzFindMt.c | 868 ++++++++++++------- deps/LZMA-SDK/C/LzFindMt.h | 36 +- deps/LZMA-SDK/C/LzFindOpt.c | 578 +++++++++++++ deps/LZMA-SDK/C/LzmaEnc.c | 351 +++++--- deps/LZMA-SDK/C/MtCoder.c | 13 +- deps/LZMA-SDK/C/MtDec.c | 8 +- deps/LZMA-SDK/C/Threads.c | 48 +- deps/LZMA-SDK/C/Threads.h | 45 +- deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.c | 7 +- deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsp | 4 + deps/LZMA-SDK/C/Util/Lzma/makefile | 2 + deps/LZMA-SDK/C/Util/Lzma/makefile.gcc | 2 + deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsp | 4 + deps/LZMA-SDK/C/Util/LzmaLib/makefile | 2 + deps/LZMA-SDK/C/XzDec.c | 5 +- deps/LZMA-SDK/C/XzIn.c | 5 +- deps/LZMA-SDK/C/var_clang_x64.mak | 1 - deps/LZMA-SDK/C/var_clang_x86.mak | 1 - deps/LZMA-SDK/C/var_gcc_x86.mak | 1 - deps/LZMA-SDK/C/warn_gcc.mak | 2 - deps/LZMA-SDK/DOC/lzma-history.txt | 36 + deps/LZMA-SDK/DOC/lzma-sdk.txt | 63 +- docs/changes.txt | 1 + 34 files changed, 2536 insertions(+), 844 deletions(-) create mode 100644 deps/LZMA-SDK/C/LzFindOpt.c diff --git a/deps/LZMA-SDK/C/7zTypes.h b/deps/LZMA-SDK/C/7zTypes.h index 497b14506..763d3a9b6 100644 --- a/deps/LZMA-SDK/C/7zTypes.h +++ b/deps/LZMA-SDK/C/7zTypes.h @@ -1,5 +1,5 @@ /* 7zTypes.h -- Basic types -2021-04-25 : Igor Pavlov : Public domain */ +2022-04-01 : Igor Pavlov : Public domain */ #ifndef __7Z_TYPES_H #define __7Z_TYPES_H @@ -62,6 +62,8 @@ typedef int SRes; typedef unsigned WRes; #define MY_SRes_HRESULT_FROM_WRes(x) HRESULT_FROM_WIN32(x) +// #define MY_HRES_ERROR__INTERNAL_ERROR MY_SRes_HRESULT_FROM_WRes(ERROR_INTERNAL_ERROR) + #else // _WIN32 // #define ENV_HAVE_LSTAT @@ -95,6 +97,7 @@ typedef int WRes; #define ERROR_DIRECTORY 267L #define ERROR_TOO_MANY_POSTS 298L +#define ERROR_INTERNAL_ERROR 1359L #define ERROR_INVALID_REPARSE_DATA 4392L #define ERROR_REPARSE_TAG_INVALID 4393L #define ERROR_REPARSE_TAG_MISMATCH 4394L @@ -102,6 +105,7 @@ typedef int WRes; // we use errno equivalents for some WIN32 errors: +#define ERROR_INVALID_PARAMETER EINVAL #define ERROR_INVALID_FUNCTION EINVAL #define ERROR_ALREADY_EXISTS EEXIST #define ERROR_FILE_EXISTS EEXIST @@ -129,10 +133,6 @@ typedef int WRes; #define MY__E_ERROR_NEGATIVE_SEEK MY_HRESULT_FROM_errno_CONST_ERROR(EINVAL) */ -// gcc / clang : (sizeof(long) == sizeof(void*)) in 32/64 bits -typedef long INT_PTR; -typedef unsigned long UINT_PTR; - #define TEXT(quote) quote #define FILE_ATTRIBUTE_READONLY 0x0001 @@ -206,6 +206,8 @@ typedef size_t SIZE_T; #endif // _WIN32 +#define MY_HRES_ERROR__INTERNAL_ERROR ((HRESULT)0x8007054FL) + #ifdef _SZ_NO_INT_64 @@ -514,6 +516,14 @@ struct ISzAlloc #endif +#define k_PropVar_TimePrec_0 0 +#define k_PropVar_TimePrec_Unix 1 +#define k_PropVar_TimePrec_DOS 2 +#define k_PropVar_TimePrec_HighPrec 3 +#define k_PropVar_TimePrec_Base 16 +#define k_PropVar_TimePrec_100ns (k_PropVar_TimePrec_Base + 7) +#define k_PropVar_TimePrec_1ns (k_PropVar_TimePrec_Base + 9) + EXTERN_C_END #endif diff --git a/deps/LZMA-SDK/C/7zVersion.h b/deps/LZMA-SDK/C/7zVersion.h index 0fe636abc..fa9e6fc53 100644 --- a/deps/LZMA-SDK/C/7zVersion.h +++ b/deps/LZMA-SDK/C/7zVersion.h @@ -1,7 +1,7 @@ -#define MY_VER_MAJOR 21 -#define MY_VER_MINOR 02 +#define MY_VER_MAJOR 22 +#define MY_VER_MINOR 01 #define MY_VER_BUILD 0 -#define MY_VERSION_NUMBERS "21.02 alpha" +#define MY_VERSION_NUMBERS "22.01" #define MY_VERSION MY_VERSION_NUMBERS #ifdef MY_CPU_NAME @@ -10,12 +10,12 @@ #define MY_VERSION_CPU MY_VERSION #endif -#define MY_DATE "2021-05-06" +#define MY_DATE "2022-07-15" #undef MY_COPYRIGHT #undef MY_VERSION_COPYRIGHT_DATE #define MY_AUTHOR_NAME "Igor Pavlov" #define MY_COPYRIGHT_PD "Igor Pavlov : Public domain" -#define MY_COPYRIGHT_CR "Copyright (c) 1999-2021 Igor Pavlov" +#define MY_COPYRIGHT_CR "Copyright (c) 1999-2022 Igor Pavlov" #ifdef USE_COPYRIGHT_CR #define MY_COPYRIGHT MY_COPYRIGHT_CR diff --git a/deps/LZMA-SDK/C/7zip_gcc_c.mak b/deps/LZMA-SDK/C/7zip_gcc_c.mak index 00ecfb043..d41810478 100644 --- a/deps/LZMA-SDK/C/7zip_gcc_c.mak +++ b/deps/LZMA-SDK/C/7zip_gcc_c.mak @@ -5,6 +5,7 @@ MY_ASM = jwasm MY_ASM = asmc PROGPATH = $(O)/$(PROG) +PROGPATH_STATIC = $(O)/$(PROG)s # for object file @@ -15,12 +16,32 @@ CFLAGS_BASE = $(MY_ARCH_2) -O2 $(CFLAGS_BASE_LIST) -Wall -Werror -Wextra $(CFLAG -DNDEBUG -D_REENTRANT -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -LDFLAGS_STATIC = -DNDEBUG -# -static - ifdef SystemDrive IS_MINGW = 1 +else +ifdef SYSTEMDRIVE +# ifdef OS +IS_MINGW = 1 endif +endif + +ifdef IS_MINGW +LDFLAGS_STATIC_2 = -static +else +ifndef DEF_FILE +ifndef IS_NOT_STANDALONE +ifndef MY_DYNAMIC_LINK +ifneq ($(CC), clang) +LDFLAGS_STATIC_2 = +# -static +# -static-libstdc++ -static-libgcc +endif +endif +endif +endif +endif + +LDFLAGS_STATIC = -DNDEBUG $(LDFLAGS_STATIC_2) ifdef DEF_FILE @@ -53,7 +74,7 @@ endif PROGPATH = $(O)/$(PROG)$(SHARED_EXT) - +PROGPATH_STATIC = $(O)/$(PROG)s$(SHARED_EXT) ifndef O O=_o @@ -61,15 +82,22 @@ endif ifdef IS_MINGW +ifdef MSYSTEM +RM = rm -f +MY_MKDIR=mkdir -p +DEL_OBJ_EXE = -$(RM) $(PROGPATH) $(PROGPATH_STATIC) $(OBJS) +else RM = del MY_MKDIR=mkdir -LIB2 = -loleaut32 -luuid -ladvapi32 -lUser32 +DEL_OBJ_EXE = -$(RM) $(O)\*.o $(O)\$(PROG).exe $(O)\$(PROG).dll +endif +LIB2 = -lOle32 -loleaut32 -luuid -ladvapi32 -lUser32 + CXXFLAGS_EXTRA = -DUNICODE -D_UNICODE # -Wno-delete-non-virtual-dtor -DEL_OBJ_EXE = -$(RM) $(O)\*.o $(O)\$(PROG).exe $(O)\$(PROG).dll else @@ -82,7 +110,7 @@ MY_MKDIR=mkdir -p # LOCAL_LIBS_DLL=$(LOCAL_LIBS) -ldl LIB2 = -lpthread -ldl -DEL_OBJ_EXE = -$(RM) $(PROGPATH) $(OBJS) +DEL_OBJ_EXE = -$(RM) $(PROGPATH) $(PROGPATH_STATIC) $(OBJS) endif @@ -108,14 +136,23 @@ CXX_WARN_FLAGS = CXXFLAGS = $(LOCAL_FLAGS) $(CXXFLAGS_BASE2) $(CFLAGS_BASE) $(CXXFLAGS_EXTRA) $(CC_SHARED) -o $@ $(CXX_WARN_FLAGS) -all: $(O) $(PROGPATH) +STATIC_TARGET= +ifdef COMPL_STATIC +STATIC_TARGET=$(PROGPATH_STATIC) +endif + + +all: $(O) $(PROGPATH) $(STATIC_TARGET) $(O): $(MY_MKDIR) $(O) +LFLAGS_ALL = -s $(MY_ARCH_2) $(LDFLAGS) $(LD_arch) $(OBJS) $(MY_LIBS) $(LIB2) $(PROGPATH): $(OBJS) - $(CXX) -s -o $(PROGPATH) $(MY_ARCH_2) $(LDFLAGS) $(OBJS) $(MY_LIBS) $(LIB2) + $(CXX) -o $(PROGPATH) $(LFLAGS_ALL) +$(PROGPATH_STATIC): $(OBJS) + $(CXX) -static -o $(PROGPATH_STATIC) $(LFLAGS_ALL) ifndef NO_DEFAULT_RES @@ -174,6 +211,8 @@ $O/LzFind.o: ../../../C/LzFind.c # ifdef MT_FILES $O/LzFindMt.o: ../../../C/LzFindMt.c $(CC) $(CFLAGS) $< +$O/LzFindOpt.o: ../../../C/LzFindOpt.c + $(CC) $(CFLAGS) $< $O/Threads.o: ../../../C/Threads.c $(CC) $(CFLAGS) $< @@ -294,7 +333,10 @@ $O/7zMain.o: ../../../C/Util/7z/7zMain.c $(CC) $(CFLAGS) $< $O/LzmaUtil.o: ../../../C/Util/Lzma/LzmaUtil.c $(CC) $(CFLAGS) $< - +$O/7zipInstall.o: ../../../C/Util/7zipInstall/7zipInstall.c + $(CC) $(CFLAGS) $< +$O/7zipUninstall.o: ../../../C/Util/7zipUninstall/7zipUninstall.c + $(CC) $(CFLAGS) $< clean: diff --git a/deps/LZMA-SDK/C/Aes.c b/deps/LZMA-SDK/C/Aes.c index 0f0ddc87a..9ad66c5c8 100644 --- a/deps/LZMA-SDK/C/Aes.c +++ b/deps/LZMA-SDK/C/Aes.c @@ -1,5 +1,5 @@ /* Aes.c -- AES encryption / decryption -2021-04-01 : Igor Pavlov : Public domain */ +2021-05-13 : Igor Pavlov : Public domain */ #include "Precomp.h" @@ -365,10 +365,10 @@ void MY_FAST_CALL AesCtr_Code(UInt32 *p, Byte *data, size_t numBlocks) #ifdef MY_CPU_LE_UNALIGN *((UInt32 *)(void *)data) ^= t; #else - data[0] ^= (t & 0xFF); - data[1] ^= ((t >> 8) & 0xFF); - data[2] ^= ((t >> 16) & 0xFF); - data[3] ^= ((t >> 24)); + data[0] = (Byte)(data[0] ^ (t & 0xFF)); + data[1] = (Byte)(data[1] ^ ((t >> 8) & 0xFF)); + data[2] = (Byte)(data[2] ^ ((t >> 16) & 0xFF)); + data[3] = (Byte)(data[3] ^ ((t >> 24))); #endif } } diff --git a/deps/LZMA-SDK/C/Alloc.c b/deps/LZMA-SDK/C/Alloc.c index 064701a8c..142a1ea22 100644 --- a/deps/LZMA-SDK/C/Alloc.c +++ b/deps/LZMA-SDK/C/Alloc.c @@ -1,12 +1,12 @@ /* Alloc.c -- Memory allocation functions -2020-10-29 : Igor Pavlov : Public domain */ +2021-07-13 : Igor Pavlov : Public domain */ #include "Precomp.h" #include #ifdef _WIN32 -#include +#include #endif #include @@ -247,14 +247,14 @@ static void *SzAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return MyAlloc static void SzFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); MyFree(address); } const ISzAlloc g_Alloc = { SzAlloc, SzFree }; +#ifdef _WIN32 static void *SzMidAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return MidAlloc(size); } static void SzMidFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); MidFree(address); } -const ISzAlloc g_MidAlloc = { SzMidAlloc, SzMidFree }; - static void *SzBigAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return BigAlloc(size); } static void SzBigFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); BigFree(address); } +const ISzAlloc g_MidAlloc = { SzMidAlloc, SzMidFree }; const ISzAlloc g_BigAlloc = { SzBigAlloc, SzBigFree }; - +#endif /* uintptr_t : C99 (optional) diff --git a/deps/LZMA-SDK/C/Alloc.h b/deps/LZMA-SDK/C/Alloc.h index a1bbe942c..59de10760 100644 --- a/deps/LZMA-SDK/C/Alloc.h +++ b/deps/LZMA-SDK/C/Alloc.h @@ -1,5 +1,5 @@ /* Alloc.h -- Memory allocation functions -2021-02-08 : Igor Pavlov : Public domain */ +2021-07-13 : Igor Pavlov : Public domain */ #ifndef __COMMON_ALLOC_H #define __COMMON_ALLOC_H @@ -30,8 +30,15 @@ void BigFree(void *address); #endif extern const ISzAlloc g_Alloc; + +#ifdef _WIN32 extern const ISzAlloc g_BigAlloc; extern const ISzAlloc g_MidAlloc; +#else +#define g_BigAlloc g_AlignedAlloc +#define g_MidAlloc g_AlignedAlloc +#endif + extern const ISzAlloc g_AlignedAlloc; diff --git a/deps/LZMA-SDK/C/CpuArch.c b/deps/LZMA-SDK/C/CpuArch.c index e1443f51b..a0e93e8b0 100644 --- a/deps/LZMA-SDK/C/CpuArch.c +++ b/deps/LZMA-SDK/C/CpuArch.c @@ -1,5 +1,5 @@ /* CpuArch.c -- CPU specific code -2021-04-28 : Igor Pavlov : Public domain */ +2021-07-13 : Igor Pavlov : Public domain */ #include "Precomp.h" @@ -217,7 +217,7 @@ BoolInt CPU_Is_InOrder() } #if !defined(MY_CPU_AMD64) && defined(_WIN32) -#include +#include static BoolInt CPU_Sys_Is_SSE_Supported() { OSVERSIONINFO vi; @@ -275,9 +275,33 @@ BoolInt CPU_IsSupported_SHA() // #include #ifdef _WIN32 -#include +#include #endif +BoolInt CPU_IsSupported_AVX2() +{ + Cx86cpuid p; + CHECK_SYS_SSE_SUPPORT + + #ifdef _WIN32 + #define MY__PF_XSAVE_ENABLED 17 + if (!IsProcessorFeaturePresent(MY__PF_XSAVE_ENABLED)) + return False; + #endif + + if (!x86cpuid_CheckAndRead(&p)) + return False; + if (p.maxFunc < 7) + return False; + { + UInt32 d[4] = { 0 }; + MyCPUID(7, &d[0], &d[1], &d[2], &d[3]); + // printf("\ncpuid(7): ebx=%8x ecx=%8x\n", d[1], d[2]); + return 1 + & (d[1] >> 5); // avx2 + } +} + BoolInt CPU_IsSupported_VAES_AVX2() { Cx86cpuid p; @@ -327,12 +351,11 @@ BoolInt CPU_IsSupported_PageGB() #ifdef _WIN32 -#include +#include -BoolInt CPU_IsSupported_CRC32() - { return IsProcessorFeaturePresent(PF_ARM_V8_CRC32_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } -BoolInt CPU_IsSupported_CRYPTO() - { return IsProcessorFeaturePresent(PF_ARM_V8_CRYPTO_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } +BoolInt CPU_IsSupported_CRC32() { return IsProcessorFeaturePresent(PF_ARM_V8_CRC32_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } +BoolInt CPU_IsSupported_CRYPTO() { return IsProcessorFeaturePresent(PF_ARM_V8_CRYPTO_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } +BoolInt CPU_IsSupported_NEON() { return IsProcessorFeaturePresent(PF_ARM_NEON_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } #else @@ -356,17 +379,27 @@ static void Print_sysctlbyname(const char *name) } */ -BoolInt CPU_IsSupported_CRC32(void) +static BoolInt My_sysctlbyname_Get_BoolInt(const char *name) { + UInt32 val = 0; + if (My_sysctlbyname_Get_UInt32(name, &val) == 0 && val == 1) + return 1; + return 0; +} + /* Print_sysctlbyname("hw.pagesize"); Print_sysctlbyname("machdep.cpu.brand_string"); */ - UInt32 val = 0; - if (My_sysctlbyname_Get_UInt32("hw.optional.armv8_crc32", &val) == 0 && val == 1) - return 1; - return 0; +BoolInt CPU_IsSupported_CRC32(void) +{ + return My_sysctlbyname_Get_BoolInt("hw.optional.armv8_crc32"); +} + +BoolInt CPU_IsSupported_NEON(void) +{ + return My_sysctlbyname_Get_BoolInt("hw.optional.neon"); } #ifdef MY_CPU_ARM64 @@ -390,18 +423,25 @@ BoolInt CPU_IsSupported_AES (void) { return APPLE_CRYPTO_SUPPORT_VAL; } #include + #define MY_HWCAP_CHECK_FUNC_2(name1, name2) \ + BoolInt CPU_IsSupported_ ## name1() { return (getauxval(AT_HWCAP) & (HWCAP_ ## name2)) ? 1 : 0; } + #ifdef MY_CPU_ARM64 #define MY_HWCAP_CHECK_FUNC(name) \ - BoolInt CPU_IsSupported_ ## name() { return (getauxval(AT_HWCAP) & (HWCAP_ ## name)) ? 1 : 0; } + MY_HWCAP_CHECK_FUNC_2(name, name) + MY_HWCAP_CHECK_FUNC_2(NEON, ASIMD) +// MY_HWCAP_CHECK_FUNC (ASIMD) #elif defined(MY_CPU_ARM) #define MY_HWCAP_CHECK_FUNC(name) \ BoolInt CPU_IsSupported_ ## name() { return (getauxval(AT_HWCAP2) & (HWCAP2_ ## name)) ? 1 : 0; } + MY_HWCAP_CHECK_FUNC_2(NEON, NEON) #endif #else // USE_HWCAP #define MY_HWCAP_CHECK_FUNC(name) \ BoolInt CPU_IsSupported_ ## name() { return 0; } + MY_HWCAP_CHECK_FUNC(NEON) #endif // USE_HWCAP diff --git a/deps/LZMA-SDK/C/CpuArch.h b/deps/LZMA-SDK/C/CpuArch.h index e1cde536d..ba8782714 100644 --- a/deps/LZMA-SDK/C/CpuArch.h +++ b/deps/LZMA-SDK/C/CpuArch.h @@ -1,5 +1,5 @@ /* CpuArch.h -- CPU specific code -2021-04-25 : Igor Pavlov : Public domain */ +2022-07-15 : Igor Pavlov : Public domain */ #ifndef __CPU_ARCH_H #define __CPU_ARCH_H @@ -123,12 +123,15 @@ MY_CPU_64BIT means that processor can work with 64-bit registers. #endif -#if defined(__sparc64__) - #define MY_CPU_NAME "sparc64" - #define MY_CPU_64BIT -#elif defined(__sparc__) - #define MY_CPU_NAME "sparc" - /* #define MY_CPU_32BIT */ +#if defined(__riscv) \ + || defined(__riscv__) + #if __riscv_xlen == 32 + #define MY_CPU_NAME "riscv32" + #elif __riscv_xlen == 64 + #define MY_CPU_NAME "riscv64" + #else + #define MY_CPU_NAME "riscv" + #endif #endif @@ -225,7 +228,6 @@ MY_CPU_64BIT means that processor can work with 64-bit registers. #endif #else #ifdef __xlC__ - // for XLC compiler: #define MY_CPU_pragma_pack_push_1 _Pragma("pack(1)") #define MY_CPU_pragma_pop _Pragma("pack()") #else @@ -253,8 +255,12 @@ MY_CPU_64BIT means that processor can work with 64-bit registers. #ifdef MY_CPU_LE #if defined(MY_CPU_X86_OR_AMD64) \ - || defined(MY_CPU_ARM64) \ - || defined(__ARM_FEATURE_UNALIGNED) + || defined(MY_CPU_ARM64) + #define MY_CPU_LE_UNALIGN + #define MY_CPU_LE_UNALIGN_64 + #elif defined(__ARM_FEATURE_UNALIGNED) + /* gcc9 for 32-bit arm can use LDRD instruction that requires 32-bit alignment. + So we can't use unaligned 64-bit operations. */ #define MY_CPU_LE_UNALIGN #endif #endif @@ -264,11 +270,15 @@ MY_CPU_64BIT means that processor can work with 64-bit registers. #define GetUi16(p) (*(const UInt16 *)(const void *)(p)) #define GetUi32(p) (*(const UInt32 *)(const void *)(p)) +#ifdef MY_CPU_LE_UNALIGN_64 #define GetUi64(p) (*(const UInt64 *)(const void *)(p)) +#endif #define SetUi16(p, v) { *(UInt16 *)(void *)(p) = (v); } #define SetUi32(p, v) { *(UInt32 *)(void *)(p) = (v); } +#ifdef MY_CPU_LE_UNALIGN_64 #define SetUi64(p, v) { *(UInt64 *)(void *)(p) = (v); } +#endif #else @@ -282,8 +292,6 @@ MY_CPU_64BIT means that processor can work with 64-bit registers. ((UInt32)((const Byte *)(p))[2] << 16) | \ ((UInt32)((const Byte *)(p))[3] << 24)) -#define GetUi64(p) (GetUi32(p) | ((UInt64)GetUi32(((const Byte *)(p)) + 4) << 32)) - #define SetUi16(p, v) { Byte *_ppp_ = (Byte *)(p); UInt32 _vvv_ = (v); \ _ppp_[0] = (Byte)_vvv_; \ _ppp_[1] = (Byte)(_vvv_ >> 8); } @@ -294,12 +302,22 @@ MY_CPU_64BIT means that processor can work with 64-bit registers. _ppp_[2] = (Byte)(_vvv_ >> 16); \ _ppp_[3] = (Byte)(_vvv_ >> 24); } +#endif + + +#ifndef MY_CPU_LE_UNALIGN_64 + +#define GetUi64(p) (GetUi32(p) | ((UInt64)GetUi32(((const Byte *)(p)) + 4) << 32)) + #define SetUi64(p, v) { Byte *_ppp2_ = (Byte *)(p); UInt64 _vvv2_ = (v); \ SetUi32(_ppp2_ , (UInt32)_vvv2_); \ SetUi32(_ppp2_ + 4, (UInt32)(_vvv2_ >> 32)); } #endif + + + #ifdef __has_builtin #define MY__has_builtin(x) __has_builtin(x) #else @@ -392,6 +410,7 @@ int x86cpuid_GetFirm(const Cx86cpuid *p); BoolInt CPU_Is_InOrder(void); BoolInt CPU_IsSupported_AES(void); +BoolInt CPU_IsSupported_AVX2(void); BoolInt CPU_IsSupported_VAES_AVX2(void); BoolInt CPU_IsSupported_SSSE3(void); BoolInt CPU_IsSupported_SSE41(void); @@ -401,6 +420,7 @@ BoolInt CPU_IsSupported_PageGB(void); #elif defined(MY_CPU_ARM_OR_ARM64) BoolInt CPU_IsSupported_CRC32(void); +BoolInt CPU_IsSupported_NEON(void); #if defined(_WIN32) BoolInt CPU_IsSupported_CRYPTO(void); diff --git a/deps/LZMA-SDK/C/DllSecur.c b/deps/LZMA-SDK/C/DllSecur.c index 19a22a9f0..a37c1b3e2 100644 --- a/deps/LZMA-SDK/C/DllSecur.c +++ b/deps/LZMA-SDK/C/DllSecur.c @@ -1,16 +1,20 @@ /* DllSecur.c -- DLL loading security -2018-02-21 : Igor Pavlov : Public domain */ +2022-07-15 : Igor Pavlov : Public domain */ #include "Precomp.h" #ifdef _WIN32 -#include +#include #include "DllSecur.h" #ifndef UNDER_CE +#if defined(__GNUC__) && (__GNUC__ >= 8) + #pragma GCC diagnostic ignored "-Wcast-function-type" +#endif + typedef BOOL (WINAPI *Func_SetDefaultDllDirectories)(DWORD DirectoryFlags); #define MY_LOAD_LIBRARY_SEARCH_USER_DIRS 0x400 @@ -33,17 +37,19 @@ static const char * const g_Dlls = #endif +// #define MY_CAST_FUNC (void(*)()) +#define MY_CAST_FUNC + void My_SetDefaultDllDirectories() { #ifndef UNDER_CE OSVERSIONINFO vi; vi.dwOSVersionInfoSize = sizeof(vi); - GetVersionEx(&vi); if (!GetVersionEx(&vi) || vi.dwMajorVersion != 6 || vi.dwMinorVersion != 0) { Func_SetDefaultDllDirectories setDllDirs = (Func_SetDefaultDllDirectories) - GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "SetDefaultDllDirectories"); + MY_CAST_FUNC GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "SetDefaultDllDirectories"); if (setDllDirs) if (setDllDirs(MY_LOAD_LIBRARY_SEARCH_SYSTEM32 | MY_LOAD_LIBRARY_SEARCH_USER_DIRS)) return; @@ -66,7 +72,7 @@ void LoadSecurityDlls() if (!GetVersionEx(&vi) || vi.dwMajorVersion != 6 || vi.dwMinorVersion != 0) { Func_SetDefaultDllDirectories setDllDirs = (Func_SetDefaultDllDirectories) - GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "SetDefaultDllDirectories"); + MY_CAST_FUNC GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "SetDefaultDllDirectories"); if (setDllDirs) if (setDllDirs(MY_LOAD_LIBRARY_SEARCH_SYSTEM32 | MY_LOAD_LIBRARY_SEARCH_USER_DIRS)) return; diff --git a/deps/LZMA-SDK/C/LzFind.c b/deps/LZMA-SDK/C/LzFind.c index 18ec00ef5..a17c06bc3 100644 --- a/deps/LZMA-SDK/C/LzFind.c +++ b/deps/LZMA-SDK/C/LzFind.c @@ -1,19 +1,29 @@ /* LzFind.c -- Match finder for LZ algorithms -2021-04-01 : Igor Pavlov : Public domain */ +2021-11-29 : Igor Pavlov : Public domain */ #include "Precomp.h" #include +// #include #include "CpuArch.h" #include "LzFind.h" #include "LzHash.h" +#define kBlockMoveAlign (1 << 7) // alignment for memmove() +#define kBlockSizeAlign (1 << 16) // alignment for block allocation +#define kBlockSizeReserveMin (1 << 24) // it's 1/256 from 4 GB dictinary + #define kEmptyHashValue 0 -#define kMaxValForNormalize ((UInt32)0xFFFFFFFF) -#define kNormalizeStepMin (1 << 10) /* it must be power of 2 */ -#define kNormalizeMask (~(UInt32)(kNormalizeStepMin - 1)) -#define kMaxHistorySize ((UInt32)7 << 29) + +#define kMaxValForNormalize ((UInt32)0) +// #define kMaxValForNormalize ((UInt32)(1 << 20) + 0xFFF) // for debug + +// #define kNormalizeAlign (1 << 7) // alignment for speculated accesses + +#define GET_AVAIL_BYTES(p) \ + Inline_MatchFinder_GetNumAvailableBytes(p) + // #define kFix5HashSize (kHash2Size + kHash3Size + kHash4Size) #define kFix5HashSize kFix4HashSize @@ -64,46 +74,57 @@ static void LzInWindow_Free(CMatchFinder *p, ISzAllocPtr alloc) } } -/* keepSizeBefore + keepSizeAfter + keepSizeReserv must be < 4G) */ -static int LzInWindow_Create(CMatchFinder *p, UInt32 keepSizeReserv, ISzAllocPtr alloc) +static int LzInWindow_Create2(CMatchFinder *p, UInt32 blockSize, ISzAllocPtr alloc) { - UInt32 blockSize = p->keepSizeBefore + p->keepSizeAfter + keepSizeReserv; - if (p->directInput) - { - p->blockSize = blockSize; - return 1; - } + if (blockSize == 0) + return 0; if (!p->bufferBase || p->blockSize != blockSize) { + // size_t blockSizeT; LzInWindow_Free(p, alloc); p->blockSize = blockSize; - p->bufferBase = (Byte *)ISzAlloc_Alloc(alloc, (size_t)blockSize); + // blockSizeT = blockSize; + + // printf("\nblockSize = 0x%x\n", blockSize); + /* + #if defined _WIN64 + // we can allocate 4GiB, but still use UInt32 for (p->blockSize) + // we use UInt32 type for (p->blockSize), because + // we don't want to wrap over 4 GiB, + // when we use (p->streamPos - p->pos) that is UInt32. + if (blockSize >= (UInt32)0 - (UInt32)kBlockSizeAlign) + { + blockSizeT = ((size_t)1 << 32); + printf("\nchanged to blockSizeT = 4GiB\n"); + } + #endif + */ + + p->bufferBase = (Byte *)ISzAlloc_Alloc(alloc, blockSize); + // printf("\nbufferBase = %p\n", p->bufferBase); + // return 0; // for debug } return (p->bufferBase != NULL); } -static Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p) { return p->buffer; } +static const Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p) { return p->buffer; } -static UInt32 MatchFinder_GetNumAvailableBytes(CMatchFinder *p) { return p->streamPos - p->pos; } +static UInt32 MatchFinder_GetNumAvailableBytes(CMatchFinder *p) { return GET_AVAIL_BYTES(p); } -void MatchFinder_ReduceOffsets(CMatchFinder *p, UInt32 subValue) -{ - p->posLimit -= subValue; - p->pos -= subValue; - p->streamPos -= subValue; -} +MY_NO_INLINE static void MatchFinder_ReadBlock(CMatchFinder *p) { if (p->streamEndWasReached || p->result != SZ_OK) return; - /* We use (p->streamPos - p->pos) value. (p->streamPos < p->pos) is allowed. */ + /* We use (p->streamPos - p->pos) value. + (p->streamPos < p->pos) is allowed. */ if (p->directInput) { - UInt32 curSize = 0xFFFFFFFF - (p->streamPos - p->pos); + UInt32 curSize = 0xFFFFFFFF - GET_AVAIL_BYTES(p); if (curSize > p->directInputRem) curSize = (UInt32)p->directInputRem; p->directInputRem -= curSize; @@ -115,10 +136,22 @@ static void MatchFinder_ReadBlock(CMatchFinder *p) for (;;) { - Byte *dest = p->buffer + (p->streamPos - p->pos); + Byte *dest = p->buffer + GET_AVAIL_BYTES(p); size_t size = (size_t)(p->bufferBase + p->blockSize - dest); if (size == 0) + { + /* we call ReadBlock() after NeedMove() and MoveBlock(). + NeedMove() and MoveBlock() povide more than (keepSizeAfter) + to the end of (blockSize). + So we don't execute this branch in normal code flow. + We can go here, if we will call ReadBlock() before NeedMove(), MoveBlock(). + */ + // p->result = SZ_ERROR_FAIL; // we can show error here return; + } + + // #define kRead 3 + // if (size > kRead) size = kRead; // for debug p->result = ISeqInStream_Read(p->stream, dest, &size); if (p->result != SZ_OK) @@ -129,41 +162,52 @@ static void MatchFinder_ReadBlock(CMatchFinder *p) return; } p->streamPos += (UInt32)size; - if (p->streamPos - p->pos > p->keepSizeAfter) + if (GET_AVAIL_BYTES(p) > p->keepSizeAfter) return; + /* here and in another (p->keepSizeAfter) checks we keep on 1 byte more than was requested by Create() function + (GET_AVAIL_BYTES(p) >= p->keepSizeAfter) - minimal required size */ } + + // on exit: (p->result != SZ_OK || p->streamEndWasReached || GET_AVAIL_BYTES(p) > p->keepSizeAfter) } + + +MY_NO_INLINE void MatchFinder_MoveBlock(CMatchFinder *p) { + const size_t offset = (size_t)(p->buffer - p->bufferBase) - p->keepSizeBefore; + const size_t keepBefore = (offset & (kBlockMoveAlign - 1)) + p->keepSizeBefore; + p->buffer = p->bufferBase + keepBefore; memmove(p->bufferBase, - p->buffer - p->keepSizeBefore, - (size_t)(p->streamPos - p->pos) + p->keepSizeBefore); - p->buffer = p->bufferBase + p->keepSizeBefore; + p->bufferBase + (offset & ~((size_t)kBlockMoveAlign - 1)), + keepBefore + (size_t)GET_AVAIL_BYTES(p)); } +/* We call MoveBlock() before ReadBlock(). + So MoveBlock() can be wasteful operation, if the whole input data + can fit in current block even without calling MoveBlock(). + in important case where (dataSize <= historySize) + condition (p->blockSize > dataSize + p->keepSizeAfter) is met + So there is no MoveBlock() in that case case. +*/ + int MatchFinder_NeedMove(CMatchFinder *p) { if (p->directInput) return 0; - /* if (p->streamEndWasReached) return 0; */ + if (p->streamEndWasReached || p->result != SZ_OK) + return 0; return ((size_t)(p->bufferBase + p->blockSize - p->buffer) <= p->keepSizeAfter); } void MatchFinder_ReadIfRequired(CMatchFinder *p) { - if (p->streamEndWasReached) - return; - if (p->keepSizeAfter >= p->streamPos - p->pos) + if (p->keepSizeAfter >= GET_AVAIL_BYTES(p)) 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) { @@ -214,32 +258,67 @@ static CLzRef* AllocRefs(size_t num, ISzAllocPtr alloc) return (CLzRef *)ISzAlloc_Alloc(alloc, sizeInBytes); } +#if (kBlockSizeReserveMin < kBlockSizeAlign * 2) + #error Stop_Compiling_Bad_Reserve +#endif + + + +static UInt32 GetBlockSize(CMatchFinder *p, UInt32 historySize) +{ + UInt32 blockSize = (p->keepSizeBefore + p->keepSizeAfter); + /* + if (historySize > kMaxHistorySize) + return 0; + */ + // printf("\nhistorySize == 0x%x\n", historySize); + + if (p->keepSizeBefore < historySize || blockSize < p->keepSizeBefore) // if 32-bit overflow + return 0; + + { + const UInt32 kBlockSizeMax = (UInt32)0 - (UInt32)kBlockSizeAlign; + const UInt32 rem = kBlockSizeMax - blockSize; + const UInt32 reserve = (blockSize >> (blockSize < ((UInt32)1 << 30) ? 1 : 2)) + + (1 << 12) + kBlockMoveAlign + kBlockSizeAlign; // do not overflow 32-bit here + if (blockSize >= kBlockSizeMax + || rem < kBlockSizeReserveMin) // we reject settings that will be slow + return 0; + if (reserve >= rem) + blockSize = kBlockSizeMax; + else + { + blockSize += reserve; + blockSize &= ~(UInt32)(kBlockSizeAlign - 1); + } + } + // printf("\n LzFind_blockSize = %x\n", blockSize); + // printf("\n LzFind_blockSize = %d\n", blockSize >> 20); + return blockSize; +} + + int MatchFinder_Create(CMatchFinder *p, UInt32 historySize, UInt32 keepAddBufferBefore, UInt32 matchMaxLen, UInt32 keepAddBufferAfter, ISzAllocPtr alloc) { - UInt32 sizeReserv; - - if (historySize > kMaxHistorySize) - { - MatchFinder_Free(p, alloc); - return 0; - } - - sizeReserv = historySize >> 1; - if (historySize >= ((UInt32)3 << 30)) sizeReserv = historySize >> 3; - else if (historySize >= ((UInt32)2 << 30)) sizeReserv = historySize >> 2; - - sizeReserv += (keepAddBufferBefore + matchMaxLen + keepAddBufferAfter) / 2 + (1 << 19); - + /* we need one additional byte in (p->keepSizeBefore), + since we use MoveBlock() after (p->pos++) and before dictionary using */ + // keepAddBufferBefore = (UInt32)0xFFFFFFFF - (1 << 22); // for debug p->keepSizeBefore = historySize + keepAddBufferBefore + 1; - 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)) + + keepAddBufferAfter += matchMaxLen; + /* we need (p->keepSizeAfter >= p->numHashBytes) */ + if (keepAddBufferAfter < p->numHashBytes) + keepAddBufferAfter = p->numHashBytes; + // keepAddBufferAfter -= 2; // for debug + p->keepSizeAfter = keepAddBufferAfter; + + if (p->directInput) + p->blockSize = 0; + if (p->directInput || LzInWindow_Create2(p, GetBlockSize(p, historySize), alloc)) { - UInt32 newCyclicBufferSize = historySize + 1; + const UInt32 newCyclicBufferSize = historySize + 1; // do not change it UInt32 hs; p->matchMaxLen = matchMaxLen; { @@ -299,7 +378,7 @@ int MatchFinder_Create(CMatchFinder *p, UInt32 historySize, size_t numSons; p->historySize = historySize; p->hashSizeSum = hs; - p->cyclicBufferSize = newCyclicBufferSize; + p->cyclicBufferSize = newCyclicBufferSize; // it must be = (historySize + 1) numSons = newCyclicBufferSize; if (p->btMode) @@ -329,33 +408,43 @@ int MatchFinder_Create(CMatchFinder *p, UInt32 historySize, return 0; } + static void MatchFinder_SetLimits(CMatchFinder *p) { - UInt32 limit = kMaxValForNormalize - p->pos; - UInt32 limit2 = p->cyclicBufferSize - p->cyclicBufferPos; + UInt32 k; + UInt32 n = kMaxValForNormalize - p->pos; + if (n == 0) + n = (UInt32)(Int32)-1; // we allow (pos == 0) at start even with (kMaxValForNormalize == 0) - if (limit2 < limit) - limit = limit2; - limit2 = p->streamPos - p->pos; - - if (limit2 <= p->keepSizeAfter) + k = p->cyclicBufferSize - p->cyclicBufferPos; + if (k < n) + n = k; + + k = GET_AVAIL_BYTES(p); { - if (limit2 > 0) - limit2 = 1; + const UInt32 ksa = p->keepSizeAfter; + UInt32 mm = p->matchMaxLen; + if (k > ksa) + k -= ksa; // we must limit exactly to keepSizeAfter for ReadBlock + else if (k >= mm) + { + // the limitation for (p->lenLimit) update + k -= mm; // optimization : to reduce the number of checks + k++; + // k = 1; // non-optimized version : for debug + } + else + { + mm = k; + if (k != 0) + k = 1; + } + p->lenLimit = mm; } - else - limit2 -= p->keepSizeAfter; + if (k < n) + n = k; - 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; + p->posLimit = p->pos + n; } @@ -363,7 +452,7 @@ void MatchFinder_Init_LowHash(CMatchFinder *p) { size_t i; CLzRef *items = p->hash; - size_t numItems = p->fixedHashSize; + const size_t numItems = p->fixedHashSize; for (i = 0; i < numItems; i++) items[i] = kEmptyHashValue; } @@ -373,80 +462,322 @@ void MatchFinder_Init_HighHash(CMatchFinder *p) { size_t i; CLzRef *items = p->hash + p->fixedHashSize; - size_t numItems = (size_t)p->hashMask + 1; + const size_t numItems = (size_t)p->hashMask + 1; for (i = 0; i < numItems; i++) items[i] = kEmptyHashValue; } -void MatchFinder_Init_3(CMatchFinder *p, int readData) +void MatchFinder_Init_4(CMatchFinder *p) { - p->cyclicBufferPos = 0; p->buffer = p->bufferBase; - p->pos = - p->streamPos = p->cyclicBufferSize; + { + /* kEmptyHashValue = 0 (Zero) is used in hash tables as NO-VALUE marker. + the code in CMatchFinderMt expects (pos = 1) */ + p->pos = + p->streamPos = + 1; // it's smallest optimal value. do not change it + // 0; // for debug + } p->result = SZ_OK; p->streamEndWasReached = 0; - - if (readData) - MatchFinder_ReadBlock(p); - - MatchFinder_SetLimits(p); } +// (CYC_TO_POS_OFFSET == 0) is expected by some optimized code +#define CYC_TO_POS_OFFSET 0 +// #define CYC_TO_POS_OFFSET 1 // for debug + void MatchFinder_Init(CMatchFinder *p) { MatchFinder_Init_HighHash(p); MatchFinder_Init_LowHash(p); - MatchFinder_Init_3(p, True); + MatchFinder_Init_4(p); + // if (readData) + MatchFinder_ReadBlock(p); + + /* if we init (cyclicBufferPos = pos), then we can use one variable + instead of both (cyclicBufferPos) and (pos) : only before (cyclicBufferPos) wrapping */ + p->cyclicBufferPos = (p->pos - CYC_TO_POS_OFFSET); // init with relation to (pos) + // p->cyclicBufferPos = 0; // smallest value + // p->son[0] = p->son[1] = 0; // unused: we can init skipped record for speculated accesses. + MatchFinder_SetLimits(p); } - -static UInt32 MatchFinder_GetSubValue(CMatchFinder *p) + + +#ifdef MY_CPU_X86_OR_AMD64 + #if defined(__clang__) && (__clang_major__ >= 8) \ + || defined(__GNUC__) && (__GNUC__ >= 8) \ + || defined(__INTEL_COMPILER) && (__INTEL_COMPILER >= 1900) + #define USE_SATUR_SUB_128 + #define USE_AVX2 + #define ATTRIB_SSE41 __attribute__((__target__("sse4.1"))) + #define ATTRIB_AVX2 __attribute__((__target__("avx2"))) + #elif defined(_MSC_VER) + #if (_MSC_VER >= 1600) + #define USE_SATUR_SUB_128 + #if (_MSC_VER >= 1900) + #define USE_AVX2 + #include // avx + #endif + #endif + #endif + +// #elif defined(MY_CPU_ARM_OR_ARM64) +#elif defined(MY_CPU_ARM64) + + #if defined(__clang__) && (__clang_major__ >= 8) \ + || defined(__GNUC__) && (__GNUC__ >= 8) + #define USE_SATUR_SUB_128 + #ifdef MY_CPU_ARM64 + // #define ATTRIB_SSE41 __attribute__((__target__(""))) + #else + // #define ATTRIB_SSE41 __attribute__((__target__("fpu=crypto-neon-fp-armv8"))) + #endif + + #elif defined(_MSC_VER) + #if (_MSC_VER >= 1910) + #define USE_SATUR_SUB_128 + #endif + #endif + + #if defined(_MSC_VER) && defined(MY_CPU_ARM64) + #include + #else + #include + #endif + +#endif + +/* +#ifndef ATTRIB_SSE41 + #define ATTRIB_SSE41 +#endif +#ifndef ATTRIB_AVX2 + #define ATTRIB_AVX2 +#endif +*/ + +#ifdef USE_SATUR_SUB_128 + +// #define _SHOW_HW_STATUS + +#ifdef _SHOW_HW_STATUS +#include +#define _PRF(x) x +_PRF(;) +#else +#define _PRF(x) +#endif + +#ifdef MY_CPU_ARM_OR_ARM64 + +#ifdef MY_CPU_ARM64 +// #define FORCE_SATUR_SUB_128 +#endif + +typedef uint32x4_t v128; +#define SASUB_128(i) \ + *(v128 *)(void *)(items + (i) * 4) = \ + vsubq_u32(vmaxq_u32(*(const v128 *)(const void *)(items + (i) * 4), sub2), sub2); + +#else + +#include // sse4.1 + +typedef __m128i v128; +#define SASUB_128(i) \ + *(v128 *)(void *)(items + (i) * 4) = \ + _mm_sub_epi32(_mm_max_epu32(*(const v128 *)(const void *)(items + (i) * 4), sub2), sub2); // SSE 4.1 + +#endif + + + +MY_NO_INLINE +static +#ifdef ATTRIB_SSE41 +ATTRIB_SSE41 +#endif +void +MY_FAST_CALL +LzFind_SaturSub_128(UInt32 subValue, CLzRef *items, const CLzRef *lim) { - return (p->pos - p->historySize - 1) & kNormalizeMask; + v128 sub2 = + #ifdef MY_CPU_ARM_OR_ARM64 + vdupq_n_u32(subValue); + #else + _mm_set_epi32((Int32)subValue, (Int32)subValue, (Int32)subValue, (Int32)subValue); + #endif + do + { + SASUB_128(0) + SASUB_128(1) + SASUB_128(2) + SASUB_128(3) + items += 4 * 4; + } + while (items != lim); } + + +#ifdef USE_AVX2 + +#include // avx + +#define SASUB_256(i) *(__m256i *)(void *)(items + (i) * 8) = _mm256_sub_epi32(_mm256_max_epu32(*(const __m256i *)(const void *)(items + (i) * 8), sub2), sub2); // AVX2 + +MY_NO_INLINE +static +#ifdef ATTRIB_AVX2 +ATTRIB_AVX2 +#endif +void +MY_FAST_CALL +LzFind_SaturSub_256(UInt32 subValue, CLzRef *items, const CLzRef *lim) +{ + __m256i sub2 = _mm256_set_epi32( + (Int32)subValue, (Int32)subValue, (Int32)subValue, (Int32)subValue, + (Int32)subValue, (Int32)subValue, (Int32)subValue, (Int32)subValue); + do + { + SASUB_256(0) + SASUB_256(1) + items += 2 * 8; + } + while (items != lim); +} +#endif // USE_AVX2 + +#ifndef FORCE_SATUR_SUB_128 +typedef void (MY_FAST_CALL *LZFIND_SATUR_SUB_CODE_FUNC)( + UInt32 subValue, CLzRef *items, const CLzRef *lim); +static LZFIND_SATUR_SUB_CODE_FUNC g_LzFind_SaturSub; +#endif // FORCE_SATUR_SUB_128 + +#endif // USE_SATUR_SUB_128 + + +// kEmptyHashValue must be zero +// #define SASUB_32(i) v = items[i]; m = v - subValue; if (v < subValue) m = kEmptyHashValue; items[i] = m; +#define SASUB_32(i) v = items[i]; if (v < subValue) v = subValue; items[i] = v - subValue; + +#ifdef FORCE_SATUR_SUB_128 + +#define DEFAULT_SaturSub LzFind_SaturSub_128 + +#else + +#define DEFAULT_SaturSub LzFind_SaturSub_32 + +MY_NO_INLINE +static +void +MY_FAST_CALL +LzFind_SaturSub_32(UInt32 subValue, CLzRef *items, const CLzRef *lim) +{ + do + { + UInt32 v; + SASUB_32(0) + SASUB_32(1) + SASUB_32(2) + SASUB_32(3) + SASUB_32(4) + SASUB_32(5) + SASUB_32(6) + SASUB_32(7) + items += 8; + } + while (items != lim); +} + +#endif + + +MY_NO_INLINE void MatchFinder_Normalize3(UInt32 subValue, CLzRef *items, size_t numItems) { - if (numItems == 0) - return; + #define K_NORM_ALIGN_BLOCK_SIZE (1 << 6) + + CLzRef *lim; + + for (; numItems != 0 && ((unsigned)(ptrdiff_t)items & (K_NORM_ALIGN_BLOCK_SIZE - 1)) != 0; numItems--) { - const CLzRef *lim = items + numItems - 1; - for (; items < lim; items += 2) - { - UInt32 v, m; - v = items[0]; m = v - subValue; if (v < subValue) m = kEmptyHashValue; - v = items[1]; items[0] = m; m = v - subValue; if (v < subValue) m = kEmptyHashValue; - items[1] = m; + UInt32 v; + SASUB_32(0); + items++; } - if (items == lim) + { - UInt32 v, m; - v = items[0]; m = v - subValue; if (v < subValue) m = kEmptyHashValue; - items[0] = m; + #define K_NORM_ALIGN_MASK (K_NORM_ALIGN_BLOCK_SIZE / 4 - 1) + lim = items + (numItems & ~(size_t)K_NORM_ALIGN_MASK); + numItems &= K_NORM_ALIGN_MASK; + if (items != lim) + { + #if defined(USE_SATUR_SUB_128) && !defined(FORCE_SATUR_SUB_128) + if (g_LzFind_SaturSub) + g_LzFind_SaturSub(subValue, items, lim); + else + #endif + DEFAULT_SaturSub(subValue, items, lim); + } + items = lim; } + + + for (; numItems != 0; numItems--) + { + UInt32 v; + SASUB_32(0); + items++; } } -static void MatchFinder_Normalize(CMatchFinder *p) -{ - UInt32 subValue = MatchFinder_GetSubValue(p); - MatchFinder_Normalize3(subValue, p->hash, p->numRefs); - MatchFinder_ReduceOffsets(p, subValue); -} +// call MatchFinder_CheckLimits() only after (p->pos++) update + MY_NO_INLINE static void MatchFinder_CheckLimits(CMatchFinder *p) { + if (// !p->streamEndWasReached && p->result == SZ_OK && + p->keepSizeAfter == GET_AVAIL_BYTES(p)) + { + // we try to read only in exact state (p->keepSizeAfter == GET_AVAIL_BYTES(p)) + if (MatchFinder_NeedMove(p)) + MatchFinder_MoveBlock(p); + MatchFinder_ReadBlock(p); + } + if (p->pos == kMaxValForNormalize) - MatchFinder_Normalize(p); - if (!p->streamEndWasReached && p->keepSizeAfter == p->streamPos - p->pos) - MatchFinder_CheckAndMoveAndRead(p); + if (GET_AVAIL_BYTES(p) >= p->numHashBytes) // optional optimization for last bytes of data. + /* + if we disable normalization for last bytes of data, and + if (data_size == 4 GiB), we don't call wastfull normalization, + but (pos) will be wrapped over Zero (0) in that case. + And we cannot resume later to normal operation + */ + { + // MatchFinder_Normalize(p); + /* after normalization we need (p->pos >= p->historySize + 1); */ + /* we can reduce subValue to aligned value, if want to keep alignment + of (p->pos) and (p->buffer) for speculated accesses. */ + const UInt32 subValue = (p->pos - p->historySize - 1) /* & ~(UInt32)(kNormalizeAlign - 1) */; + // const UInt32 subValue = (1 << 15); // for debug + // printf("\nMatchFinder_Normalize() subValue == 0x%x\n", subValue); + size_t numSonRefs = p->cyclicBufferSize; + if (p->btMode) + numSonRefs <<= 1; + Inline_MatchFinder_ReduceOffsets(p, subValue); + MatchFinder_Normalize3(subValue, p->hash, (size_t)p->hashSizeSum + numSonRefs); + } + if (p->cyclicBufferPos == p->cyclicBufferSize) p->cyclicBufferPos = 0; + MatchFinder_SetLimits(p); } @@ -455,9 +786,9 @@ static void MatchFinder_CheckLimits(CMatchFinder *p) (lenLimit > maxLen) */ MY_FORCE_INLINE -static UInt32 * Hc_GetMatchesSpec(unsigned lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, - UInt32 _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue, - UInt32 *distances, unsigned maxLen) +static UInt32 * Hc_GetMatchesSpec(size_t lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue, + UInt32 *d, unsigned maxLen) { /* son[_cyclicBufferPos] = curMatch; @@ -465,7 +796,7 @@ static UInt32 * Hc_GetMatchesSpec(unsigned lenLimit, UInt32 curMatch, UInt32 pos { UInt32 delta = pos - curMatch; if (cutValue-- == 0 || delta >= _cyclicBufferSize) - return distances; + return d; { const Byte *pb = cur - delta; curMatch = son[_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)]; @@ -478,10 +809,10 @@ static UInt32 * Hc_GetMatchesSpec(unsigned lenLimit, UInt32 curMatch, UInt32 pos if (maxLen < len) { maxLen = len; - *distances++ = len; - *distances++ = delta - 1; + *d++ = len; + *d++ = delta - 1; if (len == lenLimit) - return distances; + return d; } } } @@ -490,9 +821,15 @@ static UInt32 * Hc_GetMatchesSpec(unsigned lenLimit, UInt32 curMatch, UInt32 pos const Byte *lim = cur + lenLimit; son[_cyclicBufferPos] = curMatch; + do { - UInt32 delta = pos - curMatch; + UInt32 delta; + + if (curMatch == 0) + break; + // if (curMatch2 >= curMatch) return NULL; + delta = pos - curMatch; if (delta >= _cyclicBufferSize) break; { @@ -506,19 +843,19 @@ static UInt32 * Hc_GetMatchesSpec(unsigned lenLimit, UInt32 curMatch, UInt32 pos { if (++c == lim) { - distances[0] = (UInt32)(lim - cur); - distances[1] = delta - 1; - return distances + 2; + d[0] = (UInt32)(lim - cur); + d[1] = delta - 1; + return d + 2; } } { - unsigned len = (unsigned)(c - cur); + const unsigned len = (unsigned)(c - cur); if (maxLen < len) { maxLen = len; - distances[0] = (UInt32)len; - distances[1] = delta - 1; - distances += 2; + d[0] = (UInt32)len; + d[1] = delta - 1; + d += 2; } } } @@ -526,31 +863,36 @@ static UInt32 * Hc_GetMatchesSpec(unsigned lenLimit, UInt32 curMatch, UInt32 pos } while (--cutValue); - return distances; + return d; } MY_FORCE_INLINE UInt32 * GetMatchesSpec1(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, - UInt32 _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue, - UInt32 *distances, UInt32 maxLen) + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue, + UInt32 *d, UInt32 maxLen) { CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); unsigned len0 = 0, len1 = 0; - for (;;) + + UInt32 cmCheck; + + // if (curMatch >= pos) { *ptr0 = *ptr1 = kEmptyHashValue; return NULL; } + + cmCheck = (UInt32)(pos - _cyclicBufferSize); + if ((UInt32)pos <= _cyclicBufferSize) + cmCheck = 0; + + if (cmCheck < curMatch) + do { - UInt32 delta = pos - curMatch; - if (cutValue-- == 0 || delta >= _cyclicBufferSize) - { - *ptr0 = *ptr1 = kEmptyHashValue; - return distances; - } + const UInt32 delta = pos - curMatch; { CLzRef *pair = son + ((size_t)(_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)) << 1); const Byte *pb = cur - delta; unsigned len = (len0 < len1 ? len0 : len1); - UInt32 pair0 = pair[0]; + const UInt32 pair0 = pair[0]; if (pb[len] == cur[len]) { if (++len != lenLimit && pb[len] == cur[len]) @@ -560,48 +902,60 @@ UInt32 * GetMatchesSpec1(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byt if (maxLen < len) { maxLen = (UInt32)len; - *distances++ = (UInt32)len; - *distances++ = delta - 1; + *d++ = (UInt32)len; + *d++ = delta - 1; if (len == lenLimit) { *ptr1 = pair0; *ptr0 = pair[1]; - return distances; + return d; } } } if (pb[len] < cur[len]) { *ptr1 = curMatch; + // const UInt32 curMatch2 = pair[1]; + // if (curMatch2 >= curMatch) { *ptr0 = *ptr1 = kEmptyHashValue; return NULL; } + // curMatch = curMatch2; + curMatch = pair[1]; ptr1 = pair + 1; - curMatch = *ptr1; len1 = len; } else { *ptr0 = curMatch; + curMatch = pair[0]; ptr0 = pair; - curMatch = *ptr0; len0 = len; } } } + while(--cutValue && cmCheck < curMatch); + + *ptr0 = *ptr1 = kEmptyHashValue; + return d; } + static void SkipMatchesSpec(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, - UInt32 _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue) + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue) { CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); unsigned len0 = 0, len1 = 0; - for (;;) + + UInt32 cmCheck; + + cmCheck = (UInt32)(pos - _cyclicBufferSize); + if ((UInt32)pos <= _cyclicBufferSize) + cmCheck = 0; + + if (// curMatch >= pos || // failure + cmCheck < curMatch) + do { - UInt32 delta = pos - curMatch; - if (cutValue-- == 0 || delta >= _cyclicBufferSize) - { - *ptr0 = *ptr1 = kEmptyHashValue; - return; - } + const UInt32 delta = pos - curMatch; { CLzRef *pair = son + ((size_t)(_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)) << 1); const Byte *pb = cur - delta; @@ -623,43 +977,62 @@ static void SkipMatchesSpec(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const if (pb[len] < cur[len]) { *ptr1 = curMatch; + curMatch = pair[1]; ptr1 = pair + 1; - curMatch = *ptr1; len1 = len; } else { *ptr0 = curMatch; + curMatch = pair[0]; ptr0 = pair; - curMatch = *ptr0; len0 = len; } } } + while(--cutValue && cmCheck < curMatch); + + *ptr0 = *ptr1 = kEmptyHashValue; + return; } + #define MOVE_POS \ ++p->cyclicBufferPos; \ p->buffer++; \ - if (++p->pos == p->posLimit) MatchFinder_CheckLimits(p); + { const UInt32 pos1 = p->pos + 1; p->pos = pos1; if (pos1 == p->posLimit) MatchFinder_CheckLimits(p); } -#define MOVE_POS_RET MOVE_POS return (UInt32)offset; +#define MOVE_POS_RET MOVE_POS return distances; -static void MatchFinder_MovePos(CMatchFinder *p) { MOVE_POS; } +MY_NO_INLINE +static void MatchFinder_MovePos(CMatchFinder *p) +{ + /* we go here at the end of stream data, when (avail < num_hash_bytes) + We don't update sons[cyclicBufferPos << btMode]. + So (sons) record will contain junk. And we cannot resume match searching + to normal operation, even if we will provide more input data in buffer. + p->sons[p->cyclicBufferPos << p->btMode] = 0; // kEmptyHashValue + if (p->btMode) + p->sons[(p->cyclicBufferPos << p->btMode) + 1] = 0; // kEmptyHashValue + */ + MOVE_POS; +} #define GET_MATCHES_HEADER2(minLen, ret_op) \ - unsigned lenLimit; UInt32 hv; const Byte *cur; UInt32 curMatch; \ + unsigned lenLimit; UInt32 hv; Byte *cur; UInt32 curMatch; \ lenLimit = (unsigned)p->lenLimit; { if (lenLimit < minLen) { MatchFinder_MovePos(p); ret_op; }} \ cur = p->buffer; -#define GET_MATCHES_HEADER(minLen) GET_MATCHES_HEADER2(minLen, return 0) -#define SKIP_HEADER(minLen) GET_MATCHES_HEADER2(minLen, continue) +#define GET_MATCHES_HEADER(minLen) GET_MATCHES_HEADER2(minLen, return distances) +#define SKIP_HEADER(minLen) do { GET_MATCHES_HEADER2(minLen, continue) -#define MF_PARAMS(p) p->pos, p->buffer, p->son, p->cyclicBufferPos, p->cyclicBufferSize, p->cutValue +#define MF_PARAMS(p) lenLimit, curMatch, p->pos, p->buffer, p->son, p->cyclicBufferPos, p->cyclicBufferSize, p->cutValue + +#define SKIP_FOOTER SkipMatchesSpec(MF_PARAMS(p)); MOVE_POS; } while (--num); #define GET_MATCHES_FOOTER_BASE(_maxLen_, func) \ - offset = (unsigned)(func((UInt32)lenLimit, curMatch, MF_PARAMS(p), \ - distances + offset, (UInt32)(_maxLen_)) - distances); MOVE_POS_RET; + distances = func(MF_PARAMS(p), \ + distances, (UInt32)_maxLen_); MOVE_POS_RET; #define GET_MATCHES_FOOTER_BT(_maxLen_) \ GET_MATCHES_FOOTER_BASE(_maxLen_, GetMatchesSpec1) @@ -667,42 +1040,45 @@ static void MatchFinder_MovePos(CMatchFinder *p) { MOVE_POS; } #define GET_MATCHES_FOOTER_HC(_maxLen_) \ GET_MATCHES_FOOTER_BASE(_maxLen_, Hc_GetMatchesSpec) -#define SKIP_FOOTER \ - SkipMatchesSpec((UInt32)lenLimit, curMatch, MF_PARAMS(p)); MOVE_POS; + #define UPDATE_maxLen { \ - ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)d2; \ + const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)d2; \ const Byte *c = cur + maxLen; \ const Byte *lim = cur + lenLimit; \ for (; c != lim; c++) if (*(c + diff) != *c) break; \ maxLen = (unsigned)(c - cur); } -static UInt32 Bt2_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +static UInt32* Bt2_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) { - unsigned offset; GET_MATCHES_HEADER(2) HASH2_CALC; curMatch = p->hash[hv]; p->hash[hv] = p->pos; - offset = 0; GET_MATCHES_FOOTER_BT(1) } -UInt32 Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +UInt32* Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) { - unsigned offset; GET_MATCHES_HEADER(3) HASH_ZIP_CALC; curMatch = p->hash[hv]; p->hash[hv] = p->pos; - offset = 0; GET_MATCHES_FOOTER_BT(2) } -static UInt32 Bt3_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) + +#define SET_mmm \ + mmm = p->cyclicBufferSize; \ + if (pos < mmm) \ + mmm = pos; + + +static UInt32* Bt3_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) { + UInt32 mmm; UInt32 h2, d2, pos; - unsigned maxLen, offset; + unsigned maxLen; UInt32 *hash; GET_MATCHES_HEADER(3) @@ -718,18 +1094,19 @@ static UInt32 Bt3_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) hash[h2] = pos; (hash + kFix3HashSize)[hv] = pos; - maxLen = 2; - offset = 0; + SET_mmm - if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + maxLen = 2; + + if (d2 < mmm && *(cur - d2) == *cur) { UPDATE_maxLen distances[0] = (UInt32)maxLen; distances[1] = d2 - 1; - offset = 2; + distances += 2; if (maxLen == lenLimit) { - SkipMatchesSpec((UInt32)lenLimit, curMatch, MF_PARAMS(p)); + SkipMatchesSpec(MF_PARAMS(p)); MOVE_POS_RET; } } @@ -738,10 +1115,11 @@ static UInt32 Bt3_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) } -static UInt32 Bt4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +static UInt32* Bt4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) { + UInt32 mmm; UInt32 h2, h3, d2, d3, pos; - unsigned maxLen, offset; + unsigned maxLen; UInt32 *hash; GET_MATCHES_HEADER(4) @@ -758,43 +1136,44 @@ static UInt32 Bt4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) (hash + kFix3HashSize)[h3] = pos; (hash + kFix4HashSize)[hv] = pos; + SET_mmm + maxLen = 3; - offset = 0; for (;;) { - if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + if (d2 < mmm && *(cur - d2) == *cur) { distances[0] = 2; distances[1] = d2 - 1; - offset = 2; + distances += 2; if (*(cur - d2 + 2) == cur[2]) { - // distances[0] = 3; + // distances[-2] = 3; } - else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + else if (d3 < mmm && *(cur - d3) == *cur) { d2 = d3; - distances[2 + 1] = d3 - 1; - offset = 4; + distances[1] = d3 - 1; + distances += 2; } else break; } - else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + else if (d3 < mmm && *(cur - d3) == *cur) { d2 = d3; distances[1] = d3 - 1; - offset = 2; + distances += 2; } else break; UPDATE_maxLen - distances[(size_t)offset - 2] = (UInt32)maxLen; + distances[-2] = (UInt32)maxLen; if (maxLen == lenLimit) { - SkipMatchesSpec((UInt32)lenLimit, curMatch, MF_PARAMS(p)); + SkipMatchesSpec(MF_PARAMS(p)); MOVE_POS_RET } break; @@ -804,9 +1183,10 @@ static UInt32 Bt4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) } -static UInt32 Bt5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +static UInt32* Bt5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) { - UInt32 h2, h3, d2, d3, maxLen, offset, pos; + UInt32 mmm; + UInt32 h2, h3, d2, d3, maxLen, pos; UInt32 *hash; GET_MATCHES_HEADER(5) @@ -826,45 +1206,46 @@ static UInt32 Bt5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) // (hash + kFix4HashSize)[h4] = pos; (hash + kFix5HashSize)[hv] = pos; + SET_mmm + maxLen = 4; - offset = 0; for (;;) { - if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + if (d2 < mmm && *(cur - d2) == *cur) { distances[0] = 2; distances[1] = d2 - 1; - offset = 2; + distances += 2; if (*(cur - d2 + 2) == cur[2]) { } - else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + else if (d3 < mmm && *(cur - d3) == *cur) { - distances[3] = d3 - 1; - offset = 4; + distances[1] = d3 - 1; + distances += 2; d2 = d3; } else break; } - else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + else if (d3 < mmm && *(cur - d3) == *cur) { distances[1] = d3 - 1; - offset = 2; + distances += 2; d2 = d3; } else break; - distances[(size_t)offset - 2] = 3; + distances[-2] = 3; if (*(cur - d2 + 3) != cur[3]) break; UPDATE_maxLen - distances[(size_t)offset - 2] = maxLen; + distances[-2] = (UInt32)maxLen; if (maxLen == lenLimit) { - SkipMatchesSpec(lenLimit, curMatch, MF_PARAMS(p)); + SkipMatchesSpec(MF_PARAMS(p)); MOVE_POS_RET; } break; @@ -874,10 +1255,11 @@ static UInt32 Bt5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) } -static UInt32 Hc4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +static UInt32* Hc4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) { + UInt32 mmm; UInt32 h2, h3, d2, d3, pos; - unsigned maxLen, offset; + unsigned maxLen; UInt32 *hash; GET_MATCHES_HEADER(4) @@ -894,40 +1276,41 @@ static UInt32 Hc4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) (hash + kFix3HashSize)[h3] = pos; (hash + kFix4HashSize)[hv] = pos; + SET_mmm + maxLen = 3; - offset = 0; for (;;) { - if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + if (d2 < mmm && *(cur - d2) == *cur) { distances[0] = 2; distances[1] = d2 - 1; - offset = 2; + distances += 2; if (*(cur - d2 + 2) == cur[2]) { - // distances[0] = 3; + // distances[-2] = 3; } - else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + else if (d3 < mmm && *(cur - d3) == *cur) { d2 = d3; - distances[2 + 1] = d3 - 1; - offset = 4; + distances[1] = d3 - 1; + distances += 2; } else break; } - else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + else if (d3 < mmm && *(cur - d3) == *cur) { d2 = d3; distances[1] = d3 - 1; - offset = 2; + distances += 2; } else break; UPDATE_maxLen - distances[(size_t)offset - 2] = (UInt32)maxLen; + distances[-2] = (UInt32)maxLen; if (maxLen == lenLimit) { p->son[p->cyclicBufferPos] = curMatch; @@ -940,9 +1323,10 @@ static UInt32 Hc4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) } -static UInt32 Hc5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +static UInt32 * Hc5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) { - UInt32 h2, h3, d2, d3, maxLen, offset, pos; + UInt32 mmm; + UInt32 h2, h3, d2, d3, maxLen, pos; UInt32 *hash; GET_MATCHES_HEADER(5) @@ -962,42 +1346,43 @@ static UInt32 Hc5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) // (hash + kFix4HashSize)[h4] = pos; (hash + kFix5HashSize)[hv] = pos; + SET_mmm + maxLen = 4; - offset = 0; for (;;) { - if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + if (d2 < mmm && *(cur - d2) == *cur) { distances[0] = 2; distances[1] = d2 - 1; - offset = 2; + distances += 2; if (*(cur - d2 + 2) == cur[2]) { } - else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + else if (d3 < mmm && *(cur - d3) == *cur) { - distances[3] = d3 - 1; - offset = 4; + distances[1] = d3 - 1; + distances += 2; d2 = d3; } else break; } - else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + else if (d3 < mmm && *(cur - d3) == *cur) { distances[1] = d3 - 1; - offset = 2; + distances += 2; d2 = d3; } else break; - distances[(size_t)offset - 2] = 3; + distances[-2] = 3; if (*(cur - d2 + 3) != cur[3]) break; UPDATE_maxLen - distances[(size_t)offset - 2] = maxLen; + distances[-2] = maxLen; if (maxLen == lenLimit) { p->son[p->cyclicBufferPos] = curMatch; @@ -1010,86 +1395,75 @@ static UInt32 Hc5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) } -UInt32 Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +UInt32* Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) { - unsigned offset; GET_MATCHES_HEADER(3) HASH_ZIP_CALC; curMatch = p->hash[hv]; p->hash[hv] = p->pos; - offset = 0; GET_MATCHES_FOOTER_HC(2) } static void Bt2_MatchFinder_Skip(CMatchFinder *p, UInt32 num) { - do + SKIP_HEADER(2) { - SKIP_HEADER(2) HASH2_CALC; curMatch = p->hash[hv]; p->hash[hv] = p->pos; - SKIP_FOOTER } - while (--num != 0); + SKIP_FOOTER } void Bt3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num) { - do + SKIP_HEADER(3) { - SKIP_HEADER(3) HASH_ZIP_CALC; curMatch = p->hash[hv]; p->hash[hv] = p->pos; - SKIP_FOOTER } - while (--num != 0); + SKIP_FOOTER } static void Bt3_MatchFinder_Skip(CMatchFinder *p, UInt32 num) { - do + SKIP_HEADER(3) { UInt32 h2; UInt32 *hash; - SKIP_HEADER(3) HASH3_CALC; hash = p->hash; curMatch = (hash + kFix3HashSize)[hv]; hash[h2] = (hash + kFix3HashSize)[hv] = p->pos; - SKIP_FOOTER } - while (--num != 0); + SKIP_FOOTER } static void Bt4_MatchFinder_Skip(CMatchFinder *p, UInt32 num) { - do + SKIP_HEADER(4) { UInt32 h2, h3; UInt32 *hash; - SKIP_HEADER(4) HASH4_CALC; hash = p->hash; curMatch = (hash + kFix4HashSize)[hv]; hash [h2] = (hash + kFix3HashSize)[h3] = (hash + kFix4HashSize)[hv] = p->pos; - SKIP_FOOTER } - while (--num != 0); + SKIP_FOOTER } static void Bt5_MatchFinder_Skip(CMatchFinder *p, UInt32 num) { - do + SKIP_HEADER(5) { UInt32 h2, h3; UInt32 *hash; - SKIP_HEADER(5) HASH5_CALC; hash = p->hash; curMatch = (hash + kFix5HashSize)[hv]; @@ -1097,66 +1471,84 @@ static void Bt5_MatchFinder_Skip(CMatchFinder *p, UInt32 num) (hash + kFix3HashSize)[h3] = // (hash + kFix4HashSize)[h4] = (hash + kFix5HashSize)[hv] = p->pos; - SKIP_FOOTER } - while (--num != 0); + SKIP_FOOTER } + +#define HC_SKIP_HEADER(minLen) \ + do { if (p->lenLimit < minLen) { MatchFinder_MovePos(p); num--; continue; } { \ + Byte *cur; \ + UInt32 *hash; \ + UInt32 *son; \ + UInt32 pos = p->pos; \ + UInt32 num2 = num; \ + /* (p->pos == p->posLimit) is not allowed here !!! */ \ + { const UInt32 rem = p->posLimit - pos; if (num2 > rem) num2 = rem; } \ + num -= num2; \ + { const UInt32 cycPos = p->cyclicBufferPos; \ + son = p->son + cycPos; \ + p->cyclicBufferPos = cycPos + num2; } \ + cur = p->buffer; \ + hash = p->hash; \ + do { \ + UInt32 curMatch; \ + UInt32 hv; + + +#define HC_SKIP_FOOTER \ + cur++; pos++; *son++ = curMatch; \ + } while (--num2); \ + p->buffer = cur; \ + p->pos = pos; \ + if (pos == p->posLimit) MatchFinder_CheckLimits(p); \ + }} while(num); \ + + static void Hc4_MatchFinder_Skip(CMatchFinder *p, UInt32 num) { - do - { + HC_SKIP_HEADER(4) + UInt32 h2, h3; - UInt32 *hash; - SKIP_HEADER(4) HASH4_CALC; - hash = p->hash; curMatch = (hash + kFix4HashSize)[hv]; hash [h2] = (hash + kFix3HashSize)[h3] = - (hash + kFix4HashSize)[hv] = p->pos; - p->son[p->cyclicBufferPos] = curMatch; - MOVE_POS - } - while (--num != 0); + (hash + kFix4HashSize)[hv] = pos; + + HC_SKIP_FOOTER } + static void Hc5_MatchFinder_Skip(CMatchFinder *p, UInt32 num) { - do - { + HC_SKIP_HEADER(5) + UInt32 h2, h3; - UInt32 *hash; - SKIP_HEADER(5) - HASH5_CALC; - hash = p->hash; + HASH5_CALC curMatch = (hash + kFix5HashSize)[hv]; hash [h2] = (hash + kFix3HashSize)[h3] = // (hash + kFix4HashSize)[h4] = - (hash + kFix5HashSize)[hv] = p->pos; - p->son[p->cyclicBufferPos] = curMatch; - MOVE_POS - } - while (--num != 0); + (hash + kFix5HashSize)[hv] = pos; + + HC_SKIP_FOOTER } void Hc3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num) { - do - { - SKIP_HEADER(3) + HC_SKIP_HEADER(3) + HASH_ZIP_CALC; - curMatch = p->hash[hv]; - p->hash[hv] = p->pos; - p->son[p->cyclicBufferPos] = curMatch; - MOVE_POS - } - while (--num != 0); + curMatch = hash[hv]; + hash[hv] = pos; + + HC_SKIP_FOOTER } -void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder *vTable) + +void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder2 *vTable) { vTable->Init = (Mf_Init_Func)MatchFinder_Init; vTable->GetNumAvailableBytes = (Mf_GetNumAvailableBytes_Func)MatchFinder_GetNumAvailableBytes; @@ -1195,3 +1587,42 @@ void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder *vTable) vTable->Skip = (Mf_Skip_Func)Bt5_MatchFinder_Skip; } } + + + +void LzFindPrepare() +{ + #ifndef FORCE_SATUR_SUB_128 + #ifdef USE_SATUR_SUB_128 + LZFIND_SATUR_SUB_CODE_FUNC f = NULL; + #ifdef MY_CPU_ARM_OR_ARM64 + { + if (CPU_IsSupported_NEON()) + { + // #pragma message ("=== LzFind NEON") + _PRF(printf("\n=== LzFind NEON\n")); + f = LzFind_SaturSub_128; + } + // f = 0; // for debug + } + #else // MY_CPU_ARM_OR_ARM64 + if (CPU_IsSupported_SSE41()) + { + // #pragma message ("=== LzFind SSE41") + _PRF(printf("\n=== LzFind SSE41\n")); + f = LzFind_SaturSub_128; + + #ifdef USE_AVX2 + if (CPU_IsSupported_AVX2()) + { + // #pragma message ("=== LzFind AVX2") + _PRF(printf("\n=== LzFind AVX2\n")); + f = LzFind_SaturSub_256; + } + #endif + } + #endif // MY_CPU_ARM_OR_ARM64 + g_LzFind_SaturSub = f; + #endif // USE_SATUR_SUB_128 + #endif // FORCE_SATUR_SUB_128 +} diff --git a/deps/LZMA-SDK/C/LzFind.h b/deps/LZMA-SDK/C/LzFind.h index 3e2248e7d..8f9fade23 100644 --- a/deps/LZMA-SDK/C/LzFind.h +++ b/deps/LZMA-SDK/C/LzFind.h @@ -1,5 +1,5 @@ /* LzFind.h -- Match finder for LZ algorithms -2021-02-09 : Igor Pavlov : Public domain */ +2021-07-13 : Igor Pavlov : Public domain */ #ifndef __LZ_FIND_H #define __LZ_FIND_H @@ -15,7 +15,7 @@ typedef struct _CMatchFinder Byte *buffer; UInt32 pos; UInt32 posLimit; - UInt32 streamPos; + UInt32 streamPos; /* wrap over Zero is allowed (streamPos < pos). Use (UInt32)(streamPos - pos) */ UInt32 lenLimit; UInt32 cyclicBufferPos; @@ -51,17 +51,19 @@ typedef struct _CMatchFinder UInt64 expectedDataSize; } CMatchFinder; -#define Inline_MatchFinder_GetPointerToCurrentPos(p) ((p)->buffer) +#define Inline_MatchFinder_GetPointerToCurrentPos(p) ((const Byte *)(p)->buffer) -#define Inline_MatchFinder_GetNumAvailableBytes(p) ((p)->streamPos - (p)->pos) +#define Inline_MatchFinder_GetNumAvailableBytes(p) ((UInt32)((p)->streamPos - (p)->pos)) +/* #define Inline_MatchFinder_IsFinishedOK(p) \ ((p)->streamEndWasReached \ && (p)->streamPos == (p)->pos \ && (!(p)->directInput || (p)->directInputRem == 0)) +*/ int MatchFinder_NeedMove(CMatchFinder *p); -// Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p); +/* Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p); */ void MatchFinder_MoveBlock(CMatchFinder *p); void MatchFinder_ReadIfRequired(CMatchFinder *p); @@ -76,10 +78,21 @@ int MatchFinder_Create(CMatchFinder *p, UInt32 historySize, ISzAllocPtr alloc); void MatchFinder_Free(CMatchFinder *p, ISzAllocPtr alloc); void MatchFinder_Normalize3(UInt32 subValue, CLzRef *items, size_t numItems); -void MatchFinder_ReduceOffsets(CMatchFinder *p, UInt32 subValue); +// void MatchFinder_ReduceOffsets(CMatchFinder *p, UInt32 subValue); + +/* +#define Inline_MatchFinder_InitPos(p, val) \ + (p)->pos = (val); \ + (p)->streamPos = (val); +*/ + +#define Inline_MatchFinder_ReduceOffsets(p, subValue) \ + (p)->pos -= (subValue); \ + (p)->streamPos -= (subValue); + UInt32 * GetMatchesSpec1(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *buffer, CLzRef *son, - UInt32 _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 _cutValue, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 _cutValue, UInt32 *distances, UInt32 maxLen); /* @@ -91,7 +104,7 @@ Conditions: typedef void (*Mf_Init_Func)(void *object); typedef UInt32 (*Mf_GetNumAvailableBytes_Func)(void *object); typedef const Byte * (*Mf_GetPointerToCurrentPos_Func)(void *object); -typedef UInt32 (*Mf_GetMatches_Func)(void *object, UInt32 *distances); +typedef UInt32 * (*Mf_GetMatches_Func)(void *object, UInt32 *distances); typedef void (*Mf_Skip_Func)(void *object, UInt32); typedef struct _IMatchFinder @@ -101,21 +114,23 @@ typedef struct _IMatchFinder Mf_GetPointerToCurrentPos_Func GetPointerToCurrentPos; Mf_GetMatches_Func GetMatches; Mf_Skip_Func Skip; -} IMatchFinder; +} IMatchFinder2; -void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder *vTable); +void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder2 *vTable); void MatchFinder_Init_LowHash(CMatchFinder *p); void MatchFinder_Init_HighHash(CMatchFinder *p); -void MatchFinder_Init_3(CMatchFinder *p, int readData); +void MatchFinder_Init_4(CMatchFinder *p); void MatchFinder_Init(CMatchFinder *p); -UInt32 Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances); -UInt32 Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances); +UInt32* Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances); +UInt32* Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances); void Bt3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num); void Hc3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num); +void LzFindPrepare(void); + EXTERN_C_END #endif diff --git a/deps/LZMA-SDK/C/LzFindMt.c b/deps/LZMA-SDK/C/LzFindMt.c index cb29a1eac..3865263b7 100644 --- a/deps/LZMA-SDK/C/LzFindMt.c +++ b/deps/LZMA-SDK/C/LzFindMt.c @@ -1,8 +1,10 @@ /* LzFindMt.c -- multithreaded Match finder for LZ algorithms -2021-04-01 : Igor Pavlov : Public domain */ +2021-12-21 : Igor Pavlov : Public domain */ #include "Precomp.h" +// #include + #include "CpuArch.h" #include "LzHash.h" @@ -10,22 +12,34 @@ // #define LOG_ITERS +// #define LOG_THREAD + +#ifdef LOG_THREAD +#include +#define PRF(x) x +#else +#define PRF(x) +#endif + #ifdef LOG_ITERS #include -static UInt64 g_NumIters_Tree; -static UInt64 g_NumIters_Loop; +extern UInt64 g_NumIters_Tree; +extern UInt64 g_NumIters_Loop; +extern UInt64 g_NumIters_Bytes; #define LOG_ITER(x) x #else #define LOG_ITER(x) #endif -#define kMtHashBlockSize (1 << 17) +#define kMtHashBlockSize ((UInt32)1 << 17) #define kMtHashNumBlocks (1 << 1) -#define kMtHashNumBlocksMask (kMtHashNumBlocks - 1) -#define kMtBtBlockSize (1 << 16) +#define GET_HASH_BLOCK_OFFSET(i) (((i) & (kMtHashNumBlocks - 1)) * kMtHashBlockSize) + +#define kMtBtBlockSize ((UInt32)1 << 16) #define kMtBtNumBlocks (1 << 4) -#define kMtBtNumBlocksMask (kMtBtNumBlocks - 1) + +#define GET_BT_BLOCK_OFFSET(i) (((i) & (kMtBtNumBlocks - 1)) * (size_t)kMtBtBlockSize) /* HASH functions: @@ -36,11 +50,17 @@ static UInt64 g_NumIters_Loop; (crc[0...0xFF] & 0xFF) <-> [0...0xFF] */ +#define MF(mt) ((mt)->MatchFinder) +#define MF_CRC (p->crc) + +// #define MF(mt) (&(mt)->MatchFinder) +// #define MF_CRC (p->MatchFinder.crc) + #define MT_HASH2_CALC \ - h2 = (p->crc[cur[0]] ^ cur[1]) & (kHash2Size - 1); + h2 = (MF_CRC[cur[0]] ^ cur[1]) & (kHash2Size - 1); #define MT_HASH3_CALC { \ - UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + UInt32 temp = MF_CRC[cur[0]] ^ cur[1]; \ h2 = temp & (kHash2Size - 1); \ h3 = (temp ^ ((UInt32)cur[2] << 8)) & (kHash3Size - 1); } @@ -59,93 +79,137 @@ static UInt64 g_NumIters_Loop; */ +MY_NO_INLINE static void MtSync_Construct(CMtSync *p) { + p->affinity = 0; 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); - p->affinity = 0; } +#define DEBUG_BUFFER_LOCK // define it to debug lock state + +#ifdef DEBUG_BUFFER_LOCK +#include +#define BUFFER_MUST_BE_LOCKED(p) if (!(p)->csWasEntered) exit(1); +#define BUFFER_MUST_BE_UNLOCKED(p) if ( (p)->csWasEntered) exit(1); +#else +#define BUFFER_MUST_BE_LOCKED(p) +#define BUFFER_MUST_BE_UNLOCKED(p) +#endif + +#define LOCK_BUFFER(p) { \ + BUFFER_MUST_BE_UNLOCKED(p); \ + CriticalSection_Enter(&(p)->cs); \ + (p)->csWasEntered = True; } + +#define UNLOCK_BUFFER(p) { \ + BUFFER_MUST_BE_LOCKED(p); \ + CriticalSection_Leave(&(p)->cs); \ + (p)->csWasEntered = False; } + + MY_NO_INLINE -static void MtSync_GetNextBlock(CMtSync *p) +static UInt32 MtSync_GetNextBlock(CMtSync *p) { + UInt32 numBlocks = 0; if (p->needStart) { + BUFFER_MUST_BE_UNLOCKED(p) 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); - - // if (mt) MatchFinder_Init_LowHash(mt->MatchFinder); } else { - CriticalSection_Leave(&p->cs); - p->csWasEntered = False; - p->numProcessedBlocks++; + UNLOCK_BUFFER(p) + // we free current block + numBlocks = p->numProcessedBlocks++; Semaphore_Release1(&p->freeSemaphore); } + + // buffer is UNLOCKED here Semaphore_Wait(&p->filledSemaphore); - CriticalSection_Enter(&p->cs); - p->csWasEntered = True; + LOCK_BUFFER(p); + return numBlocks; } -/* MtSync_StopWriting must be called if Writing was started */ +/* if Writing (Processing) thread was started, we must call MtSync_StopWriting() */ + +MY_NO_INLINE static void MtSync_StopWriting(CMtSync *p) { - UInt32 myNumBlocks = p->numProcessedBlocks; if (!Thread_WasCreated(&p->thread) || p->needStart) return; - p->stopWriting = True; + + PRF(printf("\nMtSync_StopWriting %p\n", p)); + if (p->csWasEntered) { - CriticalSection_Leave(&p->cs); - p->csWasEntered = False; + /* we don't use buffer in this thread after StopWriting(). + So we UNLOCK buffer. + And we restore default UNLOCKED state for stopped thread */ + UNLOCK_BUFFER(p) } - Semaphore_Release1(&p->freeSemaphore); - - Event_Wait(&p->wasStopped); - while (myNumBlocks++ != p->numProcessedBlocks) - { - Semaphore_Wait(&p->filledSemaphore); - Semaphore_Release1(&p->freeSemaphore); - } + /* We send (p->stopWriting) message and release freeSemaphore + to free current block. + So the thread will see (p->stopWriting) at some + iteration after Wait(freeSemaphore). + The thread doesn't need to fill all avail free blocks, + so we can get fast thread stop. + */ + + p->stopWriting = True; + Semaphore_Release1(&p->freeSemaphore); // check semaphore count !!! + + PRF(printf("\nMtSync_StopWriting %p : Event_Wait(&p->wasStopped)\n", p)); + Event_Wait(&p->wasStopped); + PRF(printf("\nMtSync_StopWriting %p : Event_Wait() finsihed\n", p)); + + /* 21.03 : we don't restore samaphore counters here. + We will recreate and reinit samaphores in next start */ + p->needStart = True; } + +MY_NO_INLINE static void MtSync_Destruct(CMtSync *p) { + PRF(printf("\nMtSync_Destruct %p\n", p)); + if (Thread_WasCreated(&p->thread)) { + /* we want thread to be in Stopped state before sending EXIT command. + note: stop(btSync) will stop (htSync) also */ MtSync_StopWriting(p); + /* thread in Stopped state here : (p->needStart == true) */ p->exit = True; - if (p->needStart) - Event_Set(&p->canStart); - Thread_Wait_Close(&p->thread); + // if (p->needStart) // it's (true) + Event_Set(&p->canStart); // we send EXIT command to thread + Thread_Wait_Close(&p->thread); // we wait thread finishing } + if (p->csWasInitialized) { CriticalSection_Delete(&p->cs); p->csWasInitialized = False; } + p->csWasEntered = False; Event_Close(&p->canStart); - Event_Close(&p->wasStarted); Event_Close(&p->wasStopped); Semaphore_Close(&p->freeSemaphore); Semaphore_Close(&p->filledSemaphore); @@ -153,48 +217,75 @@ static void MtSync_Destruct(CMtSync *p) p->wasCreated = False; } -#define RINOK_THREAD(x) { if ((x) != 0) return SZ_ERROR_THREAD; } -static SRes MtSync_Create2(CMtSync *p, THREAD_FUNC_TYPE startAddress, void *obj, UInt32 numBlocks) +// #define RINOK_THREAD(x) { if ((x) != 0) return SZ_ERROR_THREAD; } +// we want to get real system error codes here instead of SZ_ERROR_THREAD +#define RINOK_THREAD(x) RINOK(x) + + +// call it before each new file (when new starting is required): +MY_NO_INLINE +static SRes MtSync_Init(CMtSync *p, UInt32 numBlocks) { WRes wres; + // BUFFER_MUST_BE_UNLOCKED(p) + if (!p->needStart || p->csWasEntered) + return SZ_ERROR_FAIL; + wres = Semaphore_OptCreateInit(&p->freeSemaphore, numBlocks, numBlocks); + if (wres == 0) + wres = Semaphore_OptCreateInit(&p->filledSemaphore, 0, numBlocks); + return MY_SRes_HRESULT_FROM_WRes(wres); +} + + +static WRes MtSync_Create_WRes(CMtSync *p, THREAD_FUNC_TYPE startAddress, void *obj) +{ + WRes wres; + if (p->wasCreated) return SZ_OK; RINOK_THREAD(CriticalSection_Init(&p->cs)); p->csWasInitialized = True; + p->csWasEntered = False; 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; + p->exit = True; /* p->exit is unused before (canStart) Event. + But in case of some unexpected code failure we will get fast exit from thread */ + + // return ERROR_TOO_MANY_POSTS; // for debug + // return EINVAL; // for debug if (p->affinity != 0) wres = Thread_Create_With_Affinity(&p->thread, startAddress, obj, (CAffinityMask)p->affinity); else wres = Thread_Create(&p->thread, startAddress, obj); + RINOK_THREAD(wres); p->wasCreated = True; return SZ_OK; } -static SRes MtSync_Create(CMtSync *p, THREAD_FUNC_TYPE startAddress, void *obj, UInt32 numBlocks) + +MY_NO_INLINE +static SRes MtSync_Create(CMtSync *p, THREAD_FUNC_TYPE startAddress, void *obj) { - SRes res = MtSync_Create2(p, startAddress, obj, numBlocks); - if (res != SZ_OK) - MtSync_Destruct(p); - return res; + const WRes wres = MtSync_Create_WRes(p, startAddress, obj); + if (wres == 0) + return 0; + MtSync_Destruct(p); + return MY_SRes_HRESULT_FROM_WRes(wres); } -// static void MtSync_Init(CMtSync *p) { p->needStart = True; } + +// ---------- HASH THREAD ---------- #define kMtMaxValForNormalize 0xFFFFFFFF -// #define kMtMaxValForNormalize ((1 << 25) + (1 << 20)) - +// #define kMtMaxValForNormalize ((1 << 21)) // for debug +// #define kNormalizeAlign (1 << 7) // alignment for speculated accesses #ifdef MY_CPU_LE_UNALIGN #define GetUi24hi_from32(p) ((UInt32)GetUi32(p) >> 8) @@ -349,27 +440,28 @@ DEF_GetHeads(5b, (crc[p[0]] ^ (crc[p[4]] << kLzHash_CrcShift_1) ^ GetUi24hi_from static void HashThreadFunc(CMatchFinderMt *mt) { CMtSync *p = &mt->hashSync; + PRF(printf("\nHashThreadFunc\n")); + for (;;) { - UInt32 numProcessedBlocks = 0; + UInt32 blockIndex = 0; + PRF(printf("\nHashThreadFunc : Event_Wait(&p->canStart)\n")); Event_Wait(&p->canStart); - Event_Set(&p->wasStarted); + PRF(printf("\nHashThreadFunc : Event_Wait(&p->canStart) : after \n")); + if (p->exit) + { + PRF(printf("\nHashThreadFunc : exit \n")); + return; + } - MatchFinder_Init_HighHash(mt->MatchFinder); + MatchFinder_Init_HighHash(MF(mt)); for (;;) { - if (p->exit) - return; - if (p->stopWriting) - { - p->numProcessedBlocks = numProcessedBlocks; - Event_Set(&p->wasStopped); - break; - } + PRF(printf("Hash thread block = %d pos = %d\n", (unsigned)blockIndex, mt->MatchFinder->pos)); { - CMatchFinder *mf = mt->MatchFinder; + CMatchFinder *mf = MF(mt); if (MatchFinder_NeedMove(mf)) { CriticalSection_Enter(&mt->btSync.cs); @@ -382,196 +474,178 @@ static void HashThreadFunc(CMatchFinderMt *mt) mt->pointerToCurPos -= offset; mt->buffer -= offset; } - CriticalSection_Leave(&mt->btSync.cs); CriticalSection_Leave(&mt->hashSync.cs); + CriticalSection_Leave(&mt->btSync.cs); continue; } Semaphore_Wait(&p->freeSemaphore); + if (p->exit) // exit is unexpected here. But we check it here for some failure case + return; + + // for faster stop : we check (p->stopWriting) after Wait(freeSemaphore) + if (p->stopWriting) + break; + MatchFinder_ReadIfRequired(mf); - if (mf->pos > (kMtMaxValForNormalize - kMtHashBlockSize)) { - UInt32 subValue = (mf->pos - mf->historySize - 1); - MatchFinder_ReduceOffsets(mf, subValue); - MatchFinder_Normalize3(subValue, mf->hash + mf->fixedHashSize, (size_t)mf->hashMask + 1); - } - { - UInt32 *heads = mt->hashBuf + ((numProcessedBlocks++) & kMtHashNumBlocksMask) * kMtHashBlockSize; - UInt32 num = mf->streamPos - mf->pos; + UInt32 *heads = mt->hashBuf + GET_HASH_BLOCK_OFFSET(blockIndex++); + UInt32 num = Inline_MatchFinder_GetNumAvailableBytes(mf); heads[0] = 2; heads[1] = num; + + /* heads[1] contains the number of avail bytes: + if (avail < mf->numHashBytes) : + { + it means that stream was finished + HASH_THREAD and BT_TREAD must move position for heads[1] (avail) bytes. + HASH_THREAD doesn't stop, + HASH_THREAD fills only the header (2 numbers) for all next blocks: + {2, NumHashBytes - 1}, {2,0}, {2,0}, ... , {2,0} + } + else + { + HASH_THREAD and BT_TREAD must move position for (heads[0] - 2) bytes; + } + */ + 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); + + if (mf->pos > (UInt32)kMtMaxValForNormalize - num) + { + const UInt32 subValue = (mf->pos - mf->historySize - 1); // & ~(UInt32)(kNormalizeAlign - 1); + Inline_MatchFinder_ReduceOffsets(mf, subValue); + MatchFinder_Normalize3(subValue, mf->hash + mf->fixedHashSize, (size_t)mf->hashMask + 1); + } + heads[0] = 2 + num; + mt->GetHeadsFunc(mf->buffer, mf->pos, mf->hash + mf->fixedHashSize, mf->hashMask, heads + 2, num, mf->crc); } - mf->pos += num; + + mf->pos += num; // wrap over zero is allowed at the end of stream mf->buffer += num; } } Semaphore_Release1(&p->filledSemaphore); - } - } + } // for() processing end + + // p->numBlocks_Sent = blockIndex; + Event_Set(&p->wasStopped); + } // for() thread end } -static void MatchFinderMt_GetNextBlock_Hash(CMatchFinderMt *p) -{ - MtSync_GetNextBlock(&p->hashSync); - p->hashBufPosLimit = p->hashBufPos = ((p->hashSync.numProcessedBlocks - 1) & kMtHashNumBlocksMask) * kMtHashBlockSize; - p->hashBufPosLimit += p->hashBuf[p->hashBufPos++]; - p->hashNumAvail = p->hashBuf[p->hashBufPos++]; -} -#define kEmptyHashValue 0 + + +// ---------- BT THREAD ---------- + +/* we use one variable instead of two (cyclicBufferPos == pos) before CyclicBuf wrap. + here we define fixed offset of (p->pos) from (p->cyclicBufferPos) */ +#define CYC_TO_POS_OFFSET 0 +// #define CYC_TO_POS_OFFSET 1 // for debug #define MFMT_GM_INLINE #ifdef MFMT_GM_INLINE /* - we use size_t for _cyclicBufferPos instead of UInt32 + we use size_t for (pos) instead of UInt32 to eliminate "movsx" BUG in old MSVC x64 compiler. */ -MY_NO_INLINE -static UInt32 *GetMatchesSpecN(UInt32 lenLimit, UInt32 pos, const Byte *cur, CLzRef *son, - size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 _cutValue, - UInt32 *d, UInt32 _maxLen, const UInt32 *hash, const UInt32 *limit, UInt32 size, UInt32 *posRes) -{ - do - { - UInt32 *_distances = ++d; - UInt32 delta = *hash++; - CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; - CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); - unsigned len0 = 0, len1 = 0; - UInt32 cutValue = _cutValue; - unsigned maxLen = (unsigned)_maxLen; - - /* - #define PREF_STEP 1 - if (size > PREF_STEP) - { - UInt32 delta = hash[PREF_STEP - 1]; - if (delta < _cyclicBufferSize) - { - size_t cyc1 = _cyclicBufferPos + PREF_STEP; - CLzRef *pair = son + ((size_t)(cyc1 - delta + ((delta > cyc1) ? _cyclicBufferSize : 0)) << 1); - Byte b = *(cur + PREF_STEP - delta); - _distances[0] = pair[0]; - _distances[1] = b; - } - } - */ - if (cutValue == 0 || delta >= _cyclicBufferSize) - { - *ptr0 = *ptr1 = kEmptyHashValue; - } - else - for (LOG_ITER(g_NumIters_Tree++);;) - { - LOG_ITER(g_NumIters_Loop++); - { - CLzRef *pair = son + ((size_t)(_cyclicBufferPos - delta + ((_cyclicBufferPos < delta) ? _cyclicBufferSize : 0)) << 1); - const Byte *pb = cur - delta; - unsigned len = (len0 < len1 ? len0 : len1); - UInt32 pair0 = *pair; - if (pb[len] == cur[len]) - { - if (++len != lenLimit && pb[len] == cur[len]) - while (++len != lenLimit) - if (pb[len] != cur[len]) - break; - if (maxLen < len) - { - maxLen = len; - *d++ = (UInt32)len; - *d++ = delta - 1; - if (len == lenLimit) - { - UInt32 pair1 = pair[1]; - *ptr1 = pair0; - *ptr0 = pair1; - break; - } - } - } - { - UInt32 curMatch = pos - delta; - // delta = pos - *pair; - // delta = pos - pair[((UInt32)pb[len] - (UInt32)cur[len]) >> 31]; - if (pb[len] < cur[len]) - { - delta = pos - pair[1]; - *ptr1 = curMatch; - ptr1 = pair + 1; - len1 = len; - } - else - { - delta = pos - *pair; - *ptr0 = curMatch; - ptr0 = pair; - len0 = len; - } - } - } - if (--cutValue == 0 || delta >= _cyclicBufferSize) - { - *ptr0 = *ptr1 = kEmptyHashValue; - break; - } - } - pos++; - _cyclicBufferPos++; - cur++; - { - UInt32 num = (UInt32)(d - _distances); - _distances[-1] = num; - } - } - while (d < limit && --size != 0); - *posRes = pos; - return d; -} +UInt32 * MY_FAST_CALL GetMatchesSpecN_2(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, + UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, + UInt32 *posRes); #endif - static void BtGetMatches(CMatchFinderMt *p, UInt32 *d) { UInt32 numProcessed = 0; UInt32 curPos = 2; - UInt32 limit = kMtBtBlockSize - (p->matchMaxLen * 2); // * 2 + /* GetMatchesSpec() functions don't create (len = 1) + in [len, dist] match pairs, if (p->numHashBytes >= 2) + Also we suppose here that (matchMaxLen >= 2). + So the following code for (reserve) is not required + UInt32 reserve = (p->matchMaxLen * 2); + const UInt32 kNumHashBytes_Max = 5; // BT_HASH_BYTES_MAX + if (reserve < kNumHashBytes_Max - 1) + reserve = kNumHashBytes_Max - 1; + const UInt32 limit = kMtBtBlockSize - (reserve); + */ + + const UInt32 limit = kMtBtBlockSize - (p->matchMaxLen * 2); + d[1] = p->hashNumAvail; + + if (p->failure_BT) + { + // printf("\n == 1 BtGetMatches() p->failure_BT\n"); + d[0] = 0; + // d[1] = 0; + return; + } while (curPos < limit) { if (p->hashBufPos == p->hashBufPosLimit) { - MatchFinderMt_GetNextBlock_Hash(p); - d[1] = numProcessed + p->hashNumAvail; - if (p->hashNumAvail >= p->numHashBytes) + // MatchFinderMt_GetNextBlock_Hash(p); + UInt32 avail; + { + const UInt32 bi = MtSync_GetNextBlock(&p->hashSync); + const UInt32 k = GET_HASH_BLOCK_OFFSET(bi); + const UInt32 *h = p->hashBuf + k; + avail = h[1]; + p->hashBufPosLimit = k + h[0]; + p->hashNumAvail = avail; + p->hashBufPos = k + 2; + } + + { + /* we must prevent UInt32 overflow for avail total value, + if avail was increased with new hash block */ + UInt32 availSum = numProcessed + avail; + if (availSum < numProcessed) + availSum = (UInt32)(Int32)-1; + d[1] = availSum; + } + + if (avail >= p->numHashBytes) continue; - d[0] = curPos + p->hashNumAvail; + + // if (p->hashBufPos != p->hashBufPosLimit) exit(1); + + /* (avail < p->numHashBytes) + It means that stream was finished. + And (avail) - is a number of remaining bytes, + we fill (d) for (avail) bytes for LZ_THREAD (receiver). + but we don't update (p->pos) and (p->cyclicBufferPos) here in BT_THREAD */ + + /* here we suppose that we have space enough: + (kMtBtBlockSize - curPos >= p->hashNumAvail) */ + p->hashNumAvail = 0; + d[0] = curPos + avail; d += curPos; - for (; p->hashNumAvail != 0; p->hashNumAvail--) + for (; avail != 0; avail--) *d++ = 0; return; } { UInt32 size = p->hashBufPosLimit - p->hashBufPos; - UInt32 lenLimit = p->matchMaxLen; UInt32 pos = p->pos; UInt32 cyclicBufferPos = p->cyclicBufferPos; + UInt32 lenLimit = p->matchMaxLen; if (lenLimit >= p->hashNumAvail) lenLimit = p->hashNumAvail; { @@ -583,6 +657,14 @@ static void BtGetMatches(CMatchFinderMt *p, UInt32 *d) size = size2; } + if (pos > (UInt32)kMtMaxValForNormalize - size) + { + const UInt32 subValue = (pos - p->cyclicBufferSize); // & ~(UInt32)(kNormalizeAlign - 1); + pos -= subValue; + p->pos = pos; + MatchFinder_Normalize3(subValue, p->son, (size_t)p->cyclicBufferSize * 2); + } + #ifndef MFMT_GM_INLINE while (curPos < limit && size-- != 0) { @@ -598,21 +680,45 @@ static void BtGetMatches(CMatchFinderMt *p, UInt32 *d) } #else { - UInt32 posRes; - curPos = (UInt32)(GetMatchesSpecN(lenLimit, pos, p->buffer, p->son, cyclicBufferPos, p->cyclicBufferSize, p->cutValue, - d + curPos, p->numHashBytes - 1, p->hashBuf + p->hashBufPos, - d + limit, - size, &posRes) - d); - p->hashBufPos += posRes - pos; - cyclicBufferPos += posRes - pos; - p->buffer += posRes - pos; - pos = posRes; + UInt32 posRes = pos; + const UInt32 *d_end; + { + d_end = GetMatchesSpecN_2( + p->buffer + lenLimit - 1, + pos, p->buffer, p->son, p->cutValue, d + curPos, + p->numHashBytes - 1, p->hashBuf + p->hashBufPos, + d + limit, p->hashBuf + p->hashBufPos + size, + cyclicBufferPos, p->cyclicBufferSize, + &posRes); + } + { + if (!d_end) + { + // printf("\n == 2 BtGetMatches() p->failure_BT\n"); + // internal data failure + p->failure_BT = True; + d[0] = 0; + // d[1] = 0; + return; + } + } + curPos = (UInt32)(d_end - d); + { + const UInt32 processed = posRes - pos; + pos = posRes; + p->hashBufPos += processed; + cyclicBufferPos += processed; + p->buffer += processed; + } } #endif - numProcessed += pos - p->pos; - p->hashNumAvail -= pos - p->pos; - p->pos = pos; + { + const UInt32 processed = pos - p->pos; + numProcessed += processed; + p->hashNumAvail -= processed; + p->pos = pos; + } if (cyclicBufferPos == p->cyclicBufferSize) cyclicBufferPos = 0; p->cyclicBufferPos = cyclicBufferPos; @@ -622,31 +728,28 @@ static void BtGetMatches(CMatchFinderMt *p, UInt32 *d) d[0] = curPos; } + static void BtFillBlock(CMatchFinderMt *p, UInt32 globalBlockIndex) { CMtSync *sync = &p->hashSync; + + BUFFER_MUST_BE_UNLOCKED(sync) + if (!sync->needStart) { - CriticalSection_Enter(&sync->cs); - sync->csWasEntered = True; + LOCK_BUFFER(sync) } - BtGetMatches(p, p->btBuf + (globalBlockIndex & kMtBtNumBlocksMask) * kMtBtBlockSize); + BtGetMatches(p, p->btBuf + GET_BT_BLOCK_OFFSET(globalBlockIndex)); + + /* We suppose that we have called GetNextBlock() from start. + So buffer is LOCKED */ - if (p->pos > kMtMaxValForNormalize - kMtBtBlockSize) - { - UInt32 subValue = p->pos - p->cyclicBufferSize; - MatchFinder_Normalize3(subValue, p->son, (size_t)p->cyclicBufferSize * 2); - p->pos -= subValue; - } - - if (!sync->needStart) - { - CriticalSection_Leave(&sync->cs); - sync->csWasEntered = False; - } + UNLOCK_BUFFER(sync) } + +MY_NO_INLINE static void BtThreadFunc(CMatchFinderMt *mt) { CMtSync *p = &mt->btSync; @@ -654,25 +757,35 @@ static void BtThreadFunc(CMatchFinderMt *mt) { UInt32 blockIndex = 0; Event_Wait(&p->canStart); - Event_Set(&p->wasStarted); + for (;;) { + PRF(printf(" BT thread block = %d pos = %d\n", (unsigned)blockIndex, mt->pos)); + /* (p->exit == true) is possible after (p->canStart) at first loop iteration + and is unexpected after more Wait(freeSemaphore) iterations */ if (p->exit) return; - if (p->stopWriting) - { - p->numProcessedBlocks = blockIndex; - MtSync_StopWriting(&mt->hashSync); - Event_Set(&p->wasStopped); - break; - } + Semaphore_Wait(&p->freeSemaphore); + + // for faster stop : we check (p->stopWriting) after Wait(freeSemaphore) + if (p->stopWriting) + break; + BtFillBlock(mt, blockIndex++); + Semaphore_Release1(&p->filledSemaphore); } + + // we stop HASH_THREAD here + MtSync_StopWriting(&mt->hashSync); + + // p->numBlocks_Sent = blockIndex; + Event_Set(&p->wasStopped); } } + void MatchFinderMt_Construct(CMatchFinderMt *p) { p->hashBuf = NULL; @@ -688,24 +801,39 @@ static void MatchFinderMt_FreeMem(CMatchFinderMt *p, ISzAllocPtr alloc) void MatchFinderMt_Destruct(CMatchFinderMt *p, ISzAllocPtr alloc) { - MtSync_Destruct(&p->hashSync); + /* + HASH_THREAD can use CriticalSection(s) btSync.cs and hashSync.cs. + So we must be sure that HASH_THREAD will not use CriticalSection(s) + after deleting CriticalSection here. + + we call ReleaseStream(p) + that calls StopWriting(btSync) + that calls StopWriting(hashSync), if it's required to stop HASH_THREAD. + after StopWriting() it's safe to destruct MtSync(s) in any order */ + + MatchFinderMt_ReleaseStream(p); + MtSync_Destruct(&p->btSync); + MtSync_Destruct(&p->hashSync); LOG_ITER( - printf("\nTree %9d * %7d iter = %9d sum \n", + printf("\nTree %9d * %7d iter = %9d = sum : bytes = %9d\n", (UInt32)(g_NumIters_Tree / 1000), (UInt32)(((UInt64)g_NumIters_Loop * 1000) / (g_NumIters_Tree + 1)), - (UInt32)(g_NumIters_Loop / 1000) + (UInt32)(g_NumIters_Loop / 1000), + (UInt32)(g_NumIters_Bytes / 1000) )); MatchFinderMt_FreeMem(p, alloc); } + #define kHashBufferSize (kMtHashBlockSize * kMtHashNumBlocks) #define kBtBufferSize (kMtBtBlockSize * kMtBtNumBlocks) -static THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE HashThreadFunc2(void *p) { HashThreadFunc((CMatchFinderMt *)p); return 0; } -static THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE BtThreadFunc2(void *p) + +static THREAD_FUNC_DECL HashThreadFunc2(void *p) { HashThreadFunc((CMatchFinderMt *)p); return 0; } +static THREAD_FUNC_DECL BtThreadFunc2(void *p) { Byte allocaDummy[0x180]; unsigned i = 0; @@ -716,16 +844,17 @@ static THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE BtThreadFunc2(void *p) return 0; } + SRes MatchFinderMt_Create(CMatchFinderMt *p, UInt32 historySize, UInt32 keepAddBufferBefore, UInt32 matchMaxLen, UInt32 keepAddBufferAfter, ISzAllocPtr alloc) { - CMatchFinder *mf = p->MatchFinder; + CMatchFinder *mf = MF(p); p->historySize = historySize; if (kMtBtBlockSize <= matchMaxLen * 4) return SZ_ERROR_PARAM; if (!p->hashBuf) { - p->hashBuf = (UInt32 *)ISzAlloc_Alloc(alloc, (kHashBufferSize + kBtBufferSize) * sizeof(UInt32)); + p->hashBuf = (UInt32 *)ISzAlloc_Alloc(alloc, ((size_t)kHashBufferSize + (size_t)kBtBufferSize) * sizeof(UInt32)); if (!p->hashBuf) return SZ_ERROR_MEM; p->btBuf = p->hashBuf + kHashBufferSize; @@ -735,101 +864,163 @@ SRes MatchFinderMt_Create(CMatchFinderMt *p, UInt32 historySize, UInt32 keepAddB if (!MatchFinder_Create(mf, historySize, keepAddBufferBefore, matchMaxLen, keepAddBufferAfter, alloc)) return SZ_ERROR_MEM; - RINOK(MtSync_Create(&p->hashSync, HashThreadFunc2, p, kMtHashNumBlocks)); - RINOK(MtSync_Create(&p->btSync, BtThreadFunc2, p, kMtBtNumBlocks)); + RINOK(MtSync_Create(&p->hashSync, HashThreadFunc2, p)); + RINOK(MtSync_Create(&p->btSync, BtThreadFunc2, p)); return SZ_OK; } -/* Call it after ReleaseStream / SetStream */ + +SRes MatchFinderMt_InitMt(CMatchFinderMt *p) +{ + RINOK(MtSync_Init(&p->hashSync, kMtHashNumBlocks)); + return MtSync_Init(&p->btSync, kMtBtNumBlocks); +} + + static void MatchFinderMt_Init(CMatchFinderMt *p) { - CMatchFinder *mf = p->MatchFinder; + CMatchFinder *mf = MF(p); p->btBufPos = - p->btBufPosLimit = 0; + p->btBufPosLimit = NULL; p->hashBufPos = p->hashBufPosLimit = 0; + p->hashNumAvail = 0; // 21.03 + + p->failure_BT = False; /* Init without data reading. We don't want to read data in this thread */ - MatchFinder_Init_3(mf, False); + MatchFinder_Init_4(mf); + MatchFinder_Init_LowHash(mf); p->pointerToCurPos = Inline_MatchFinder_GetPointerToCurrentPos(mf); p->btNumAvailBytes = 0; - p->lzPos = p->historySize + 1; + p->failure_LZ_BT = False; + // p->failure_LZ_LZ = False; + + p->lzPos = + 1; // optimal smallest value + // 0; // for debug: ignores match to start + // kNormalizeAlign; // for debug p->hash = mf->hash; p->fixedHashSize = mf->fixedHashSize; // p->hash4Mask = mf->hash4Mask; p->crc = mf->crc; + // memcpy(p->crc, mf->crc, sizeof(mf->crc)); p->son = mf->son; p->matchMaxLen = mf->matchMaxLen; p->numHashBytes = mf->numHashBytes; - p->pos = mf->pos; - p->buffer = mf->buffer; - p->cyclicBufferPos = mf->cyclicBufferPos; + + /* (mf->pos) and (mf->streamPos) were already initialized to 1 in MatchFinder_Init_4() */ + // mf->streamPos = mf->pos = 1; // optimal smallest value + // 0; // for debug: ignores match to start + // kNormalizeAlign; // for debug + + /* we must init (p->pos = mf->pos) for BT, because + BT code needs (p->pos == delta_value_for_empty_hash_record == mf->pos) */ + p->pos = mf->pos; // do not change it + + p->cyclicBufferPos = (p->pos - CYC_TO_POS_OFFSET); p->cyclicBufferSize = mf->cyclicBufferSize; + p->buffer = mf->buffer; p->cutValue = mf->cutValue; + // p->son[0] = p->son[1] = 0; // unused: to init skipped record for speculated accesses. } + /* ReleaseStream is required to finish multithreading */ void MatchFinderMt_ReleaseStream(CMatchFinderMt *p) { + // Sleep(1); // for debug MtSync_StopWriting(&p->btSync); + // Sleep(200); // for debug /* p->MatchFinder->ReleaseStream(); */ } MY_NO_INLINE -static void MatchFinderMt_GetNextBlock_Bt(CMatchFinderMt *p) +static UInt32 MatchFinderMt_GetNextBlock_Bt(CMatchFinderMt *p) { - UInt32 blockIndex, k; - - MtSync_GetNextBlock(&p->btSync); - - blockIndex = ((p->btSync.numProcessedBlocks - 1) & kMtBtNumBlocksMask); - k = blockIndex * kMtBtBlockSize; - p->btBufPosLimit = k + p->btBuf[k]; - p->btNumAvailBytes = p->btBuf[k + 1]; - p->btBufPos = k + 2; - if (p->lzPos >= kMtMaxValForNormalize - kMtBtBlockSize) + if (p->failure_LZ_BT) + p->btBufPos = p->failureBuf; + else { - MatchFinder_Normalize3(p->lzPos - p->historySize - 1, p->hash, p->fixedHashSize); - p->lzPos = p->historySize + 1; + const UInt32 bi = MtSync_GetNextBlock(&p->btSync); + const UInt32 *bt = p->btBuf + GET_BT_BLOCK_OFFSET(bi); + { + const UInt32 numItems = bt[0]; + p->btBufPosLimit = bt + numItems; + p->btNumAvailBytes = bt[1]; + p->btBufPos = bt + 2; + if (numItems < 2 || numItems > kMtBtBlockSize) + { + p->failureBuf[0] = 0; + p->btBufPos = p->failureBuf; + p->btBufPosLimit = p->failureBuf + 1; + p->failure_LZ_BT = True; + // p->btNumAvailBytes = 0; + /* we don't want to decrease AvailBytes, that was load before. + that can be unxepected for the code that have loaded anopther value before */ + } + } + + if (p->lzPos >= (UInt32)kMtMaxValForNormalize - (UInt32)kMtBtBlockSize) + { + /* we don't check (lzPos) over exact avail bytes in (btBuf). + (fixedHashSize) is small, so normalization is fast */ + const UInt32 subValue = (p->lzPos - p->historySize - 1); // & ~(UInt32)(kNormalizeAlign - 1); + p->lzPos -= subValue; + MatchFinder_Normalize3(subValue, p->hash, p->fixedHashSize); + } } + return p->btNumAvailBytes; } + + static const Byte * MatchFinderMt_GetPointerToCurrentPos(CMatchFinderMt *p) { return p->pointerToCurPos; } + #define GET_NEXT_BLOCK_IF_REQUIRED if (p->btBufPos == p->btBufPosLimit) MatchFinderMt_GetNextBlock_Bt(p); + static UInt32 MatchFinderMt_GetNumAvailableBytes(CMatchFinderMt *p) { - GET_NEXT_BLOCK_IF_REQUIRED; - return p->btNumAvailBytes; + if (p->btBufPos != p->btBufPosLimit) + return p->btNumAvailBytes; + return MatchFinderMt_GetNextBlock_Bt(p); } + +// #define CHECK_FAILURE_LZ(_match_, _pos_) if (_match_ >= _pos_) { p->failure_LZ_LZ = True; return d; } +#define CHECK_FAILURE_LZ(_match_, _pos_) + static UInt32 * MixMatches2(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *d) { UInt32 h2, c2; UInt32 *hash = p->hash; const Byte *cur = p->pointerToCurPos; - UInt32 m = p->lzPos; + const UInt32 m = p->lzPos; MT_HASH2_CALC c2 = hash[h2]; hash[h2] = m; if (c2 >= matchMinPos) + { + CHECK_FAILURE_LZ(c2, m) if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m] == cur[0]) { *d++ = 2; *d++ = m - c2 - 1; } + } return d; } @@ -839,7 +1030,7 @@ static UInt32 * MixMatches3(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *d) UInt32 h2, h3, c2, c3; UInt32 *hash = p->hash; const Byte *cur = p->pointerToCurPos; - UInt32 m = p->lzPos; + const UInt32 m = p->lzPos; MT_HASH3_CALC c2 = hash[h2]; @@ -848,22 +1039,30 @@ static UInt32 * MixMatches3(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *d) hash[h2] = m; (hash + kFix3HashSize)[h3] = m; - if (c2 >= matchMinPos && cur[(ptrdiff_t)c2 - (ptrdiff_t)m] == cur[0]) + if (c2 >= matchMinPos) { - d[1] = m - c2 - 1; - if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m + 2] == cur[2]) + CHECK_FAILURE_LZ(c2, m) + if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m] == cur[0]) { - d[0] = 3; - return d + 2; + d[1] = m - c2 - 1; + if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m + 2] == cur[2]) + { + d[0] = 3; + return d + 2; + } + d[0] = 2; + d += 2; } - d[0] = 2; - d += 2; } - if (c3 >= matchMinPos && cur[(ptrdiff_t)c3 - (ptrdiff_t)m] == cur[0]) + if (c3 >= matchMinPos) { - *d++ = 3; - *d++ = m - c3 - 1; + CHECK_FAILURE_LZ(c3, m) + if (cur[(ptrdiff_t)c3 - (ptrdiff_t)m] == cur[0]) + { + *d++ = 3; + *d++ = m - c3 - 1; + } } return d; @@ -874,30 +1073,37 @@ static UInt32 * MixMatches3(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *d) /* static -UInt32 MatchFinderMt_GetMatches_Bt4(CMatchFinderMt *p, UInt32 *d) +UInt32* MatchFinderMt_GetMatches_Bt4(CMatchFinderMt *p, UInt32 *d) { - UInt32 pos = p->btBufPos; - const UInt32 *bt = p->btBuf + pos; - UInt32 len = *bt++; + const UInt32 *bt = p->btBufPos; + const UInt32 len = *bt++; + const UInt32 *btLim = bt + len; UInt32 matchMinPos; - const UInt32 *d_base = d; UInt32 avail = p->btNumAvailBytes - 1; - p->btBufPos = pos + 1 + len; + p->btBufPos = btLim; { - UInt32 temp1 = p->historySize; p->btNumAvailBytes = avail; #define BT_HASH_BYTES_MAX 5 + + matchMinPos = p->lzPos; if (len != 0) - temp1 = bt[1]; - else if (avail < (BT_HASH_BYTES_MAX - 2)) + matchMinPos -= bt[1]; + else if (avail < (BT_HASH_BYTES_MAX - 1) - 1) { INCREASE_LZ_POS - return 0; + return d; + } + else + { + const UInt32 hs = p->historySize; + if (matchMinPos > hs) + matchMinPos -= hs; + else + matchMinPos = 1; } - matchMinPos = p->lzPos - temp1; } for (;;) @@ -942,17 +1148,17 @@ UInt32 MatchFinderMt_GetMatches_Bt4(CMatchFinderMt *p, UInt32 *d) { do { - UInt32 v0 = bt[0]; - UInt32 v1 = bt[1]; + const UInt32 v0 = bt[0]; + const UInt32 v1 = bt[1]; bt += 2; d[0] = v0; d[1] = v1; d += 2; } - while ((len -= 2) != 0); + while (bt != btLim); } INCREASE_LZ_POS - return (UInt32)(d - d_base); + return d; } */ @@ -962,7 +1168,7 @@ static UInt32 *MixMatches4(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *d) UInt32 h2, h3, /* h4, */ c2, c3 /* , c4 */; UInt32 *hash = p->hash; const Byte *cur = p->pointerToCurPos; - UInt32 m = p->lzPos; + const UInt32 m = p->lzPos; MT_HASH3_CALC // MT_HASH4_CALC c2 = hash[h2]; @@ -1038,43 +1244,49 @@ static UInt32 *MixMatches4(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *d) } -static UInt32 MatchFinderMt2_GetMatches(CMatchFinderMt *p, UInt32 *d) +static UInt32* MatchFinderMt2_GetMatches(CMatchFinderMt *p, UInt32 *d) { - const UInt32 *bt = p->btBuf + p->btBufPos; - UInt32 len = *bt++; - p->btBufPos += 1 + len; + const UInt32 *bt = p->btBufPos; + const UInt32 len = *bt++; + const UInt32 *btLim = bt + len; + p->btBufPos = btLim; p->btNumAvailBytes--; + INCREASE_LZ_POS { - UInt32 i; - for (i = 0; i < len; i += 2) + while (bt != btLim) { - UInt32 v0 = bt[0]; - UInt32 v1 = bt[1]; + const UInt32 v0 = bt[0]; + const UInt32 v1 = bt[1]; bt += 2; d[0] = v0; d[1] = v1; d += 2; } } - INCREASE_LZ_POS - return len; + return d; } -static UInt32 MatchFinderMt_GetMatches(CMatchFinderMt *p, UInt32 *d) +static UInt32* MatchFinderMt_GetMatches(CMatchFinderMt *p, UInt32 *d) { - UInt32 pos = p->btBufPos; - const UInt32 *bt = p->btBuf + pos; + const UInt32 *bt = p->btBufPos; UInt32 len = *bt++; - UInt32 avail = p->btNumAvailBytes - 1; + const UInt32 avail = p->btNumAvailBytes - 1; p->btNumAvailBytes = avail; - p->btBufPos = pos + 1 + len; + p->btBufPos = bt + len; if (len == 0) { #define BT_HASH_BYTES_MAX 5 if (avail >= (BT_HASH_BYTES_MAX - 1) - 1) - len = (UInt32)(p->MixMatchesFunc(p, p->lzPos - p->historySize, d) - d); + { + UInt32 m = p->lzPos; + if (m > p->historySize) + m -= p->historySize; + else + m = 1; + d = p->MixMatchesFunc(p, m, d); + } } else { @@ -1083,27 +1295,26 @@ static UInt32 MatchFinderMt_GetMatches(CMatchFinderMt *p, UInt32 *d) (match_len >= numHashBytes). MixMatchesFunc() inserts only hash matches that are nearer than (match_dist) */ - UInt32 *d2; - d2 = p->MixMatchesFunc(p, p->lzPos - bt[1], d); + d = p->MixMatchesFunc(p, p->lzPos - bt[1], d); + // if (d) // check for failure do { - UInt32 v0 = bt[0]; - UInt32 v1 = bt[1]; + const UInt32 v0 = bt[0]; + const UInt32 v1 = bt[1]; bt += 2; - d2[0] = v0; - d2[1] = v1; - d2 += 2; + d[0] = v0; + d[1] = v1; + d += 2; } - while ((len -= 2) != 0); - len = (UInt32)(d2 - d); + while (len -= 2); } INCREASE_LZ_POS - return len; + return d; } #define SKIP_HEADER2_MT do { GET_NEXT_BLOCK_IF_REQUIRED #define SKIP_HEADER_MT(n) SKIP_HEADER2_MT if (p->btNumAvailBytes-- >= (n)) { const Byte *cur = p->pointerToCurPos; UInt32 *hash = p->hash; -#define SKIP_FOOTER_MT } INCREASE_LZ_POS p->btBufPos += p->btBuf[p->btBufPos] + 1; } while (--num != 0); +#define SKIP_FOOTER_MT } INCREASE_LZ_POS p->btBufPos += (size_t)*p->btBufPos + 1; } while (--num != 0); static void MatchFinderMt0_Skip(CMatchFinderMt *p, UInt32 num) { @@ -1131,10 +1342,14 @@ static void MatchFinderMt3_Skip(CMatchFinderMt *p, UInt32 num) SKIP_FOOTER_MT } +/* +// MatchFinderMt4_Skip() is similar to MatchFinderMt3_Skip(). +// The difference is that MatchFinderMt3_Skip() updates hash for last 3 bytes of stream. + static void MatchFinderMt4_Skip(CMatchFinderMt *p, UInt32 num) { SKIP_HEADER_MT(4) - UInt32 h2, h3 /*, h4 */; + UInt32 h2, h3; // h4 MT_HASH3_CALC // MT_HASH4_CALC // (hash + kFix4HashSize)[h4] = @@ -1143,15 +1358,16 @@ static void MatchFinderMt4_Skip(CMatchFinderMt *p, UInt32 num) p->lzPos; SKIP_FOOTER_MT } +*/ -void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder *vTable) +void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder2 *vTable) { vTable->Init = (Mf_Init_Func)MatchFinderMt_Init; vTable->GetNumAvailableBytes = (Mf_GetNumAvailableBytes_Func)MatchFinderMt_GetNumAvailableBytes; vTable->GetPointerToCurrentPos = (Mf_GetPointerToCurrentPos_Func)MatchFinderMt_GetPointerToCurrentPos; vTable->GetMatches = (Mf_GetMatches_Func)MatchFinderMt_GetMatches; - switch (p->MatchFinder->numHashBytes) + switch (MF(p)->numHashBytes) { case 2: p->GetHeadsFunc = GetHeads2; @@ -1160,12 +1376,12 @@ void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder *vTable) vTable->GetMatches = (Mf_GetMatches_Func)MatchFinderMt2_GetMatches; break; case 3: - p->GetHeadsFunc = p->MatchFinder->bigHash ? GetHeads3b : GetHeads3; + p->GetHeadsFunc = MF(p)->bigHash ? GetHeads3b : GetHeads3; p->MixMatchesFunc = (Mf_Mix_Matches)MixMatches2; vTable->Skip = (Mf_Skip_Func)MatchFinderMt2_Skip; break; case 4: - p->GetHeadsFunc = p->MatchFinder->bigHash ? GetHeads4b : GetHeads4; + p->GetHeadsFunc = MF(p)->bigHash ? GetHeads4b : GetHeads4; // it's fast inline version of GetMatches() // vTable->GetMatches = (Mf_GetMatches_Func)MatchFinderMt_GetMatches_Bt4; @@ -1174,9 +1390,11 @@ void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder *vTable) vTable->Skip = (Mf_Skip_Func)MatchFinderMt3_Skip; break; default: - p->GetHeadsFunc = p->MatchFinder->bigHash ? GetHeads5b : GetHeads5; + p->GetHeadsFunc = MF(p)->bigHash ? GetHeads5b : GetHeads5; p->MixMatchesFunc = (Mf_Mix_Matches)MixMatches4; - vTable->Skip = (Mf_Skip_Func)MatchFinderMt4_Skip; + vTable->Skip = + (Mf_Skip_Func)MatchFinderMt3_Skip; + // (Mf_Skip_Func)MatchFinderMt4_Skip; break; } } diff --git a/deps/LZMA-SDK/C/LzFindMt.h b/deps/LZMA-SDK/C/LzFindMt.h index 888c787cb..ee9a1b6fe 100644 --- a/deps/LZMA-SDK/C/LzFindMt.h +++ b/deps/LZMA-SDK/C/LzFindMt.h @@ -1,5 +1,5 @@ /* LzFindMt.h -- multithreaded Match finder for LZ algorithms -2019-11-05 : Igor Pavlov : Public domain */ +2021-07-12 : Igor Pavlov : Public domain */ #ifndef __LZ_FIND_MT_H #define __LZ_FIND_MT_H @@ -11,22 +11,24 @@ EXTERN_C_BEGIN typedef struct _CMtSync { + UInt32 numProcessedBlocks; + CThread thread; + UInt64 affinity; + BoolInt wasCreated; BoolInt needStart; + BoolInt csWasInitialized; + BoolInt csWasEntered; + BoolInt exit; BoolInt stopWriting; - CThread thread; CAutoResetEvent canStart; - CAutoResetEvent wasStarted; CAutoResetEvent wasStopped; CSemaphore freeSemaphore; CSemaphore filledSemaphore; - BoolInt csWasInitialized; - BoolInt csWasEntered; CCriticalSection cs; - UInt32 numProcessedBlocks; - UInt64 affinity; + // UInt32 numBlocks_Sent; } CMtSync; typedef UInt32 * (*Mf_Mix_Matches)(void *p, UInt32 matchMinPos, UInt32 *distances); @@ -42,8 +44,8 @@ typedef struct _CMatchFinderMt /* LZ */ const Byte *pointerToCurPos; UInt32 *btBuf; - UInt32 btBufPos; - UInt32 btBufPosLimit; + const UInt32 *btBufPos; + const UInt32 *btBufPosLimit; UInt32 lzPos; UInt32 btNumAvailBytes; @@ -54,6 +56,10 @@ typedef struct _CMatchFinderMt const UInt32 *crc; Mf_Mix_Matches MixMatchesFunc; + UInt32 failure_LZ_BT; // failure in BT transfered to LZ + // UInt32 failure_LZ_LZ; // failure in LZ tables + UInt32 failureBuf[1]; + // UInt32 crc[256]; /* LZ + BT */ CMtSync btSync; @@ -64,6 +70,8 @@ typedef struct _CMatchFinderMt UInt32 hashBufPos; UInt32 hashBufPosLimit; UInt32 hashNumAvail; + UInt32 failure_BT; + CLzRef *son; UInt32 matchMaxLen; @@ -71,7 +79,7 @@ typedef struct _CMatchFinderMt UInt32 pos; const Byte *buffer; UInt32 cyclicBufferPos; - UInt32 cyclicBufferSize; /* it must be historySize + 1 */ + UInt32 cyclicBufferSize; /* it must be = (historySize + 1) */ UInt32 cutValue; /* BT + Hash */ @@ -81,13 +89,19 @@ typedef struct _CMatchFinderMt /* Hash */ Mf_GetHeads GetHeadsFunc; CMatchFinder *MatchFinder; + // CMatchFinder MatchFinder; } CMatchFinderMt; +// only for Mt part void MatchFinderMt_Construct(CMatchFinderMt *p); void MatchFinderMt_Destruct(CMatchFinderMt *p, ISzAllocPtr alloc); + SRes MatchFinderMt_Create(CMatchFinderMt *p, UInt32 historySize, UInt32 keepAddBufferBefore, UInt32 matchMaxLen, UInt32 keepAddBufferAfter, ISzAllocPtr alloc); -void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder *vTable); +void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder2 *vTable); + +/* call MatchFinderMt_InitMt() before IMatchFinder::Init() */ +SRes MatchFinderMt_InitMt(CMatchFinderMt *p); void MatchFinderMt_ReleaseStream(CMatchFinderMt *p); EXTERN_C_END diff --git a/deps/LZMA-SDK/C/LzFindOpt.c b/deps/LZMA-SDK/C/LzFindOpt.c new file mode 100644 index 000000000..dbb6ad9fc --- /dev/null +++ b/deps/LZMA-SDK/C/LzFindOpt.c @@ -0,0 +1,578 @@ +/* LzFindOpt.c -- multithreaded Match finder for LZ algorithms +2021-07-13 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" +#include "LzFind.h" + +// #include "LzFindMt.h" + +// #define LOG_ITERS + +// #define LOG_THREAD + +#ifdef LOG_THREAD +#include +#define PRF(x) x +#else +// #define PRF(x) +#endif + +#ifdef LOG_ITERS +#include +UInt64 g_NumIters_Tree; +UInt64 g_NumIters_Loop; +UInt64 g_NumIters_Bytes; +#define LOG_ITER(x) x +#else +#define LOG_ITER(x) +#endif + +// ---------- BT THREAD ---------- + +#define USE_SON_PREFETCH +#define USE_LONG_MATCH_OPT + +#define kEmptyHashValue 0 + +// #define CYC_TO_POS_OFFSET 0 + +// #define CYC_TO_POS_OFFSET 1 // for debug + +/* +MY_NO_INLINE +UInt32 * MY_FAST_CALL GetMatchesSpecN_1(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, + UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, UInt32 *posRes) +{ + do + { + UInt32 delta; + if (hash == size) + break; + delta = *hash++; + + if (delta == 0 || delta > (UInt32)pos) + return NULL; + + lenLimit++; + + if (delta == (UInt32)pos) + { + CLzRef *ptr1 = son + ((size_t)pos << 1) - CYC_TO_POS_OFFSET * 2; + *d++ = 0; + ptr1[0] = kEmptyHashValue; + ptr1[1] = kEmptyHashValue; + } +else +{ + UInt32 *_distances = ++d; + + CLzRef *ptr0 = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2 + 1; + CLzRef *ptr1 = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2; + + const Byte *len0 = cur, *len1 = cur; + UInt32 cutValue = _cutValue; + const Byte *maxLen = cur + _maxLen; + + for (LOG_ITER(g_NumIters_Tree++);;) + { + LOG_ITER(g_NumIters_Loop++); + { + const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; + CLzRef *pair = son + ((size_t)(((ptrdiff_t)pos - CYC_TO_POS_OFFSET) + diff) << 1); + const Byte *len = (len0 < len1 ? len0 : len1); + + #ifdef USE_SON_PREFETCH + const UInt32 pair0 = *pair; + #endif + + if (len[diff] == len[0]) + { + if (++len != lenLimit && len[diff] == len[0]) + while (++len != lenLimit) + { + LOG_ITER(g_NumIters_Bytes++); + if (len[diff] != len[0]) + break; + } + if (maxLen < len) + { + maxLen = len; + *d++ = (UInt32)(len - cur); + *d++ = delta - 1; + + if (len == lenLimit) + { + const UInt32 pair1 = pair[1]; + *ptr1 = + #ifdef USE_SON_PREFETCH + pair0; + #else + pair[0]; + #endif + *ptr0 = pair1; + + _distances[-1] = (UInt32)(d - _distances); + + #ifdef USE_LONG_MATCH_OPT + + if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) + break; + + { + for (;;) + { + hash++; + pos++; + cur++; + lenLimit++; + { + CLzRef *ptr = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2; + #if 0 + *(UInt64 *)(void *)ptr = ((const UInt64 *)(const void *)ptr)[diff]; + #else + const UInt32 p0 = ptr[0 + (diff * 2)]; + const UInt32 p1 = ptr[1 + (diff * 2)]; + ptr[0] = p0; + ptr[1] = p1; + // ptr[0] = ptr[0 + (diff * 2)]; + // ptr[1] = ptr[1 + (diff * 2)]; + #endif + } + // PrintSon(son + 2, pos - 1); + // printf("\npos = %x delta = %x\n", pos, delta); + len++; + *d++ = 2; + *d++ = (UInt32)(len - cur); + *d++ = delta - 1; + if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) + break; + } + } + #endif + + break; + } + } + } + + { + const UInt32 curMatch = (UInt32)pos - delta; // (UInt32)(pos + diff); + if (len[diff] < len[0]) + { + delta = pair[1]; + if (delta >= curMatch) + return NULL; + *ptr1 = curMatch; + ptr1 = pair + 1; + len1 = len; + } + else + { + delta = *pair; + if (delta >= curMatch) + return NULL; + *ptr0 = curMatch; + ptr0 = pair; + len0 = len; + } + + delta = (UInt32)pos - delta; + + if (--cutValue == 0 || delta >= pos) + { + *ptr0 = *ptr1 = kEmptyHashValue; + _distances[-1] = (UInt32)(d - _distances); + break; + } + } + } + } // for (tree iterations) +} + pos++; + cur++; + } + while (d < limit); + *posRes = (UInt32)pos; + return d; +} +*/ + +/* define cbs if you use 2 functions. + GetMatchesSpecN_1() : (pos < _cyclicBufferSize) + GetMatchesSpecN_2() : (pos >= _cyclicBufferSize) + + do not define cbs if you use 1 function: + GetMatchesSpecN_2() +*/ + +// #define cbs _cyclicBufferSize + +/* + we use size_t for (pos) and (_cyclicBufferPos_ instead of UInt32 + to eliminate "movsx" BUG in old MSVC x64 compiler. +*/ + +UInt32 * MY_FAST_CALL GetMatchesSpecN_2(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, + UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, + UInt32 *posRes); + +MY_NO_INLINE +UInt32 * MY_FAST_CALL GetMatchesSpecN_2(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, + UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, + UInt32 *posRes) +{ + do // while (hash != size) + { + UInt32 delta; + + #ifndef cbs + UInt32 cbs; + #endif + + if (hash == size) + break; + + delta = *hash++; + + if (delta == 0) + return NULL; + + lenLimit++; + + #ifndef cbs + cbs = _cyclicBufferSize; + if ((UInt32)pos < cbs) + { + if (delta > (UInt32)pos) + return NULL; + cbs = (UInt32)pos; + } + #endif + + if (delta >= cbs) + { + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + *d++ = 0; + ptr1[0] = kEmptyHashValue; + ptr1[1] = kEmptyHashValue; + } +else +{ + UInt32 *_distances = ++d; + + CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + + UInt32 cutValue = _cutValue; + const Byte *len0 = cur, *len1 = cur; + const Byte *maxLen = cur + _maxLen; + + // if (cutValue == 0) { *ptr0 = *ptr1 = kEmptyHashValue; } else + for (LOG_ITER(g_NumIters_Tree++);;) + { + LOG_ITER(g_NumIters_Loop++); + { + // SPEC code + CLzRef *pair = son + ((size_t)((ptrdiff_t)_cyclicBufferPos - (ptrdiff_t)delta + + (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0) + ) << 1); + + const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; + const Byte *len = (len0 < len1 ? len0 : len1); + + #ifdef USE_SON_PREFETCH + const UInt32 pair0 = *pair; + #endif + + if (len[diff] == len[0]) + { + if (++len != lenLimit && len[diff] == len[0]) + while (++len != lenLimit) + { + LOG_ITER(g_NumIters_Bytes++); + if (len[diff] != len[0]) + break; + } + if (maxLen < len) + { + maxLen = len; + *d++ = (UInt32)(len - cur); + *d++ = delta - 1; + + if (len == lenLimit) + { + const UInt32 pair1 = pair[1]; + *ptr1 = + #ifdef USE_SON_PREFETCH + pair0; + #else + pair[0]; + #endif + *ptr0 = pair1; + + _distances[-1] = (UInt32)(d - _distances); + + #ifdef USE_LONG_MATCH_OPT + + if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) + break; + + { + for (;;) + { + *d++ = 2; + *d++ = (UInt32)(lenLimit - cur); + *d++ = delta - 1; + cur++; + lenLimit++; + // SPEC + _cyclicBufferPos++; + { + // SPEC code + CLzRef *dest = son + ((size_t)(_cyclicBufferPos) << 1); + const CLzRef *src = dest + ((diff + + (ptrdiff_t)(UInt32)((_cyclicBufferPos < delta) ? cbs : 0)) << 1); + // CLzRef *ptr = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2; + #if 0 + *(UInt64 *)(void *)dest = *((const UInt64 *)(const void *)src); + #else + const UInt32 p0 = src[0]; + const UInt32 p1 = src[1]; + dest[0] = p0; + dest[1] = p1; + #endif + } + pos++; + hash++; + if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) + break; + } // for() end for long matches + } + #endif + + break; // break from TREE iterations + } + } + } + { + const UInt32 curMatch = (UInt32)pos - delta; // (UInt32)(pos + diff); + if (len[diff] < len[0]) + { + delta = pair[1]; + *ptr1 = curMatch; + ptr1 = pair + 1; + len1 = len; + if (delta >= curMatch) + return NULL; + } + else + { + delta = *pair; + *ptr0 = curMatch; + ptr0 = pair; + len0 = len; + if (delta >= curMatch) + return NULL; + } + delta = (UInt32)pos - delta; + + if (--cutValue == 0 || delta >= cbs) + { + *ptr0 = *ptr1 = kEmptyHashValue; + _distances[-1] = (UInt32)(d - _distances); + break; + } + } + } + } // for (tree iterations) +} + pos++; + _cyclicBufferPos++; + cur++; + } + while (d < limit); + *posRes = (UInt32)pos; + return d; +} + + + +/* +typedef UInt32 uint32plus; // size_t + +UInt32 * MY_FAST_CALL GetMatchesSpecN_3(uint32plus lenLimit, size_t pos, const Byte *cur, CLzRef *son, + UInt32 _cutValue, UInt32 *d, uint32plus _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, + UInt32 *posRes) +{ + do // while (hash != size) + { + UInt32 delta; + + #ifndef cbs + UInt32 cbs; + #endif + + if (hash == size) + break; + + delta = *hash++; + + if (delta == 0) + return NULL; + + #ifndef cbs + cbs = _cyclicBufferSize; + if ((UInt32)pos < cbs) + { + if (delta > (UInt32)pos) + return NULL; + cbs = (UInt32)pos; + } + #endif + + if (delta >= cbs) + { + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + *d++ = 0; + ptr1[0] = kEmptyHashValue; + ptr1[1] = kEmptyHashValue; + } +else +{ + CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + UInt32 *_distances = ++d; + uint32plus len0 = 0, len1 = 0; + UInt32 cutValue = _cutValue; + uint32plus maxLen = _maxLen; + // lenLimit++; // const Byte *lenLimit = cur + _lenLimit; + + for (LOG_ITER(g_NumIters_Tree++);;) + { + LOG_ITER(g_NumIters_Loop++); + { + // const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; + CLzRef *pair = son + ((size_t)((ptrdiff_t)_cyclicBufferPos - delta + + (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0) + ) << 1); + const Byte *pb = cur - delta; + uint32plus len = (len0 < len1 ? len0 : len1); + + #ifdef USE_SON_PREFETCH + const UInt32 pair0 = *pair; + #endif + + if (pb[len] == cur[len]) + { + if (++len != lenLimit && pb[len] == cur[len]) + while (++len != lenLimit) + if (pb[len] != cur[len]) + break; + if (maxLen < len) + { + maxLen = len; + *d++ = (UInt32)len; + *d++ = delta - 1; + if (len == lenLimit) + { + { + const UInt32 pair1 = pair[1]; + *ptr0 = pair1; + *ptr1 = + #ifdef USE_SON_PREFETCH + pair0; + #else + pair[0]; + #endif + } + + _distances[-1] = (UInt32)(d - _distances); + + #ifdef USE_LONG_MATCH_OPT + + if (hash == size || *hash != delta || pb[lenLimit] != cur[lenLimit] || d >= limit) + break; + + { + const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; + for (;;) + { + *d++ = 2; + *d++ = (UInt32)lenLimit; + *d++ = delta - 1; + _cyclicBufferPos++; + { + CLzRef *dest = son + ((size_t)_cyclicBufferPos << 1); + const CLzRef *src = dest + ((diff + + (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0)) << 1); + #if 0 + *(UInt64 *)(void *)dest = *((const UInt64 *)(const void *)src); + #else + const UInt32 p0 = src[0]; + const UInt32 p1 = src[1]; + dest[0] = p0; + dest[1] = p1; + #endif + } + hash++; + pos++; + cur++; + pb++; + if (hash == size || *hash != delta || pb[lenLimit] != cur[lenLimit] || d >= limit) + break; + } + } + #endif + + break; + } + } + } + { + const UInt32 curMatch = (UInt32)pos - delta; + if (pb[len] < cur[len]) + { + delta = pair[1]; + *ptr1 = curMatch; + ptr1 = pair + 1; + len1 = len; + } + else + { + delta = *pair; + *ptr0 = curMatch; + ptr0 = pair; + len0 = len; + } + + { + if (delta >= curMatch) + return NULL; + delta = (UInt32)pos - delta; + if (delta >= cbs + // delta >= _cyclicBufferSize || delta >= pos + || --cutValue == 0) + { + *ptr0 = *ptr1 = kEmptyHashValue; + _distances[-1] = (UInt32)(d - _distances); + break; + } + } + } + } + } // for (tree iterations) +} + pos++; + _cyclicBufferPos++; + cur++; + } + while (d < limit); + *posRes = (UInt32)pos; + return d; +} +*/ diff --git a/deps/LZMA-SDK/C/LzmaEnc.c b/deps/LZMA-SDK/C/LzmaEnc.c index 86dcb1963..ca9154aef 100644 --- a/deps/LZMA-SDK/C/LzmaEnc.c +++ b/deps/LZMA-SDK/C/LzmaEnc.c @@ -1,5 +1,5 @@ /* LzmaEnc.c -- LZMA Encoder -2021-04-01: Igor Pavlov : Public domain */ +2022-07-15: Igor Pavlov : Public domain */ #include "Precomp.h" @@ -12,6 +12,7 @@ #include #endif +#include "CpuArch.h" #include "LzmaEnc.h" #include "LzFind.h" @@ -36,8 +37,8 @@ void LzmaEnc_RestoreState(CLzmaEncHandle pp); static unsigned g_STAT_OFFSET = 0; #endif -#define kLzmaMaxHistorySize ((UInt32)3 << 29) -/* #define kLzmaMaxHistorySize ((UInt32)7 << 29) */ +/* for good normalization speed we still reserve 256 MB before 4 GB range */ +#define kLzmaMaxHistorySize ((UInt32)15 << 28) #define kNumTopBits 24 #define kTopValue ((UInt32)1 << kNumTopBits) @@ -78,13 +79,12 @@ void LzmaEncProps_Normalize(CLzmaEncProps *p) if (p->dictSize > p->reduceSize) { - unsigned i; - UInt32 reduceSize = (UInt32)p->reduceSize; - for (i = 11; i <= 30; i++) - { - if (reduceSize <= ((UInt32)2 << i)) { p->dictSize = ((UInt32)2 << i); break; } - if (reduceSize <= ((UInt32)3 << i)) { p->dictSize = ((UInt32)3 << i); break; } - } + UInt32 v = (UInt32)p->reduceSize; + const UInt32 kReduceMin = ((UInt32)1 << 12); + if (v < kReduceMin) + v = kReduceMin; + if (p->dictSize > v) + p->dictSize = v; } if (p->lc < 0) p->lc = 3; @@ -113,18 +113,85 @@ UInt32 LzmaEncProps_GetDictSize(const CLzmaEncProps *props2) return props.dictSize; } -#if defined(_MSC_VER) && (_MSC_VER >= 1400) -/* BSR code is fast for some new CPUs */ -/* #define LZMA_LOG_BSR */ + +/* +x86/x64: + +BSR: + IF (SRC == 0) ZF = 1, DEST is undefined; + AMD : DEST is unchanged; + IF (SRC != 0) ZF = 0; DEST is index of top non-zero bit + BSR is slow in some processors + +LZCNT: + IF (SRC == 0) CF = 1, DEST is size_in_bits_of_register(src) (32 or 64) + IF (SRC != 0) CF = 0, DEST = num_lead_zero_bits + IF (DEST == 0) ZF = 1; + +LZCNT works only in new processors starting from Haswell. +if LZCNT is not supported by processor, then it's executed as BSR. +LZCNT can be faster than BSR, if supported. +*/ + +// #define LZMA_LOG_BSR + +#if defined(MY_CPU_ARM_OR_ARM64) /* || defined(MY_CPU_X86_OR_AMD64) */ + + #if (defined(__clang__) && (__clang_major__ >= 6)) \ + || (defined(__GNUC__) && (__GNUC__ >= 6)) + #define LZMA_LOG_BSR + #elif defined(_MSC_VER) && (_MSC_VER >= 1300) + // #if defined(MY_CPU_ARM_OR_ARM64) + #define LZMA_LOG_BSR + // #endif + #endif #endif +// #include + #ifdef LZMA_LOG_BSR -#define kDicLogSizeMaxCompress 32 +#if defined(__clang__) \ + || defined(__GNUC__) -#define BSR2_RET(pos, res) { unsigned long zz; _BitScanReverse(&zz, (pos)); res = (zz + zz) + ((pos >> (zz - 1)) & 1); } +/* + C code: : (30 - __builtin_clz(x)) + gcc9/gcc10 for x64 /x86 : 30 - (bsr(x) xor 31) + clang10 for x64 : 31 + (bsr(x) xor -32) +*/ -static unsigned GetPosSlot1(UInt32 pos) + #define MY_clz(x) ((unsigned)__builtin_clz(x)) + // __lzcnt32 + // __builtin_ia32_lzcnt_u32 + +#else // #if defined(_MSC_VER) + + #ifdef MY_CPU_ARM_OR_ARM64 + + #define MY_clz _CountLeadingZeros + + #else // if defined(MY_CPU_X86_OR_AMD64) + + // #define MY_clz __lzcnt // we can use lzcnt (unsupported by old CPU) + // _BitScanReverse code is not optimal for some MSVC compilers + #define BSR2_RET(pos, res) { unsigned long zz; _BitScanReverse(&zz, (pos)); zz--; \ + res = (zz + zz) + (pos >> zz); } + + #endif // MY_CPU_X86_OR_AMD64 + +#endif // _MSC_VER + + +#ifndef BSR2_RET + + #define BSR2_RET(pos, res) { unsigned zz = 30 - MY_clz(pos); \ + res = (zz + zz) + (pos >> zz); } + +#endif + + +unsigned GetPosSlot1(UInt32 pos); +unsigned GetPosSlot1(UInt32 pos) { unsigned res; BSR2_RET(pos, res); @@ -133,10 +200,10 @@ static unsigned GetPosSlot1(UInt32 pos) #define GetPosSlot2(pos, res) { BSR2_RET(pos, res); } #define GetPosSlot(pos, res) { if (pos < 2) res = pos; else BSR2_RET(pos, res); } -#else -#define kNumLogBits (9 + sizeof(size_t) / 2) -/* #define kNumLogBits (11 + sizeof(size_t) / 8 * 3) */ +#else // ! LZMA_LOG_BSR + +#define kNumLogBits (11 + sizeof(size_t) / 8 * 3) #define kDicLogSizeMaxCompress ((kNumLogBits - 1) * 2 + 7) @@ -183,7 +250,7 @@ static void LzmaEnc_FastPosInit(Byte *g_FastPos) #define GetPosSlot2(pos, res) { BSR2_RET(pos, res); } #define GetPosSlot(pos, res) { if (pos < kNumFullDistances) res = p->g_FastPos[pos & (kNumFullDistances - 1)]; else BSR2_RET(pos, res); } -#endif +#endif // LZMA_LOG_BSR #define LZMA_NUM_REPS 4 @@ -319,7 +386,7 @@ typedef UInt32 CProbPrice; typedef struct { void *matchFinderObj; - IMatchFinder matchFinder; + IMatchFinder2 matchFinder; unsigned optCur; unsigned optEnd; @@ -364,10 +431,14 @@ typedef struct // begin of CMatchFinderMt is used in LZ thread CMatchFinderMt matchFinderMt; // end of CMatchFinderMt is used in BT and HASH threads + // #else + // CMatchFinder matchFinderBase; #endif - CMatchFinder matchFinderBase; + + // we suppose that we have 8-bytes alignment after CMatchFinder + #ifndef _7ZIP_ST Byte pad[128]; #endif @@ -375,8 +446,10 @@ typedef struct // LZ thread CProbPrice ProbPrices[kBitModelTotal >> kNumMoveReducingBits]; - UInt32 matches[LZMA_MATCH_LEN_MAX * 2 + 2 + 1]; + // we want {len , dist} pairs to be 8-bytes aligned in matches array + UInt32 matches[LZMA_MATCH_LEN_MAX * 2 + 2]; + // we want 8-bytes alignment here UInt32 alignPrices[kAlignTableSize]; UInt32 posSlotPrices[kNumLenToPosStates][kDistTableSizeMax]; UInt32 distancesPrices[kNumLenToPosStates][kNumFullDistances]; @@ -405,12 +478,19 @@ typedef struct CSaveState saveState; + // BoolInt mf_Failure; #ifndef _7ZIP_ST Byte pad2[128]; #endif } CLzmaEnc; +#define MFB (p->matchFinderBase) +/* +#ifndef _7ZIP_ST +#define MFB (p->matchFinderMt.MatchFinder) +#endif +*/ #define COPY_ARR(dest, src, arr) memcpy(dest->arr, src->arr, sizeof(src->arr)); @@ -475,11 +555,21 @@ SRes LzmaEnc_SetProps(CLzmaEncHandle pp, const CLzmaEncProps *props2) if (props.lc > LZMA_LC_MAX || props.lp > LZMA_LP_MAX - || props.pb > LZMA_PB_MAX - || props.dictSize > ((UInt64)1 << kDicLogSizeMaxCompress) - || props.dictSize > kLzmaMaxHistorySize) + || props.pb > LZMA_PB_MAX) return SZ_ERROR_PARAM; + + if (props.dictSize > kLzmaMaxHistorySize) + props.dictSize = kLzmaMaxHistorySize; + + #ifndef LZMA_LOG_BSR + { + const UInt64 dict64 = props.dictSize; + if (dict64 > ((UInt64)1 << kDicLogSizeMaxCompress)) + return SZ_ERROR_PARAM; + } + #endif + p->dictSize = props.dictSize; { unsigned fb = (unsigned)props.fb; @@ -494,7 +584,7 @@ SRes LzmaEnc_SetProps(CLzmaEncHandle pp, const CLzmaEncProps *props2) p->pb = (unsigned)props.pb; p->fastMode = (props.algo == 0); // p->_maxMode = True; - p->matchFinderBase.btMode = (Byte)(props.btMode ? 1 : 0); + MFB.btMode = (Byte)(props.btMode ? 1 : 0); { unsigned numHashBytes = 4; if (props.btMode) @@ -504,10 +594,10 @@ SRes LzmaEnc_SetProps(CLzmaEncHandle pp, const CLzmaEncProps *props2) } if (props.numHashBytes >= 5) numHashBytes = 5; - p->matchFinderBase.numHashBytes = numHashBytes; + MFB.numHashBytes = numHashBytes; } - p->matchFinderBase.cutValue = props.mc; + MFB.cutValue = props.mc; p->writeEndMark = (BoolInt)props.writeEndMark; @@ -531,7 +621,7 @@ SRes LzmaEnc_SetProps(CLzmaEncHandle pp, const CLzmaEncProps *props2) void LzmaEnc_SetDataSize(CLzmaEncHandle pp, UInt64 expectedDataSiize) { CLzmaEnc *p = (CLzmaEnc *)pp; - p->matchFinderBase.expectedDataSize = expectedDataSiize; + MFB.expectedDataSize = expectedDataSiize; } @@ -578,12 +668,11 @@ static int RangeEnc_Alloc(CRangeEnc *p, ISzAllocPtr alloc) static void RangeEnc_Free(CRangeEnc *p, ISzAllocPtr alloc) { ISzAlloc_Free(alloc, p->bufBase); - p->bufBase = 0; + p->bufBase = NULL; } static void RangeEnc_Init(CRangeEnc *p) { - /* Stream.Init(); */ p->range = 0xFFFFFFFF; p->cache = 0; p->low = 0; @@ -597,12 +686,12 @@ static void RangeEnc_Init(CRangeEnc *p) MY_NO_INLINE static void RangeEnc_FlushStream(CRangeEnc *p) { - size_t num; - if (p->res != SZ_OK) - return; - num = (size_t)(p->buf - p->bufBase); - if (num != ISeqOutStream_Write(p->outStream, p->bufBase, num)) - p->res = SZ_ERROR_WRITE; + const size_t num = (size_t)(p->buf - p->bufBase); + if (p->res == SZ_OK) + { + if (num != ISeqOutStream_Write(p->outStream, p->bufBase, num)) + p->res = SZ_ERROR_WRITE; + } p->processed += num; p->buf = p->bufBase; } @@ -1007,7 +1096,11 @@ static unsigned ReadMatchDistances(CLzmaEnc *p, unsigned *numPairsRes) p->additionalOffset++; p->numAvail = p->matchFinder.GetNumAvailableBytes(p->matchFinderObj); - numPairs = p->matchFinder.GetMatches(p->matchFinderObj, p->matches); + { + const UInt32 *d = p->matchFinder.GetMatches(p->matchFinderObj, p->matches); + // if (!d) { p->mf_Failure = True; *numPairsRes = 0; return 0; } + numPairs = (unsigned)(d - p->matches); + } *numPairsRes = numPairs; #ifdef SHOW_STAT @@ -1023,7 +1116,7 @@ static unsigned ReadMatchDistances(CLzmaEnc *p, unsigned *numPairsRes) if (numPairs == 0) return 0; { - unsigned len = p->matches[(size_t)numPairs - 2]; + const unsigned len = p->matches[(size_t)numPairs - 2]; if (len != p->numFastBytes) return len; { @@ -1033,7 +1126,7 @@ static unsigned ReadMatchDistances(CLzmaEnc *p, unsigned *numPairsRes) { const Byte *p1 = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; const Byte *p2 = p1 + len; - ptrdiff_t dif = (ptrdiff_t)-1 - (ptrdiff_t)p->matches[(size_t)numPairs - 1]; + const ptrdiff_t dif = (ptrdiff_t)-1 - (ptrdiff_t)p->matches[(size_t)numPairs - 1]; const Byte *lim = p1 + numAvail; for (; p2 != lim && *p2 == p2[dif]; p2++) {} @@ -1189,6 +1282,8 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position) repLens[i] = len; if (len > repLens[repMaxIndex]) repMaxIndex = i; + if (len == LZMA_MATCH_LEN_MAX) // 21.03 : optimization + break; } if (repLens[repMaxIndex] >= p->numFastBytes) @@ -1201,10 +1296,12 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position) } matches = p->matches; + #define MATCHES matches + // #define MATCHES p->matches if (mainLen >= p->numFastBytes) { - p->backRes = matches[(size_t)numPairs - 1] + LZMA_NUM_REPS; + p->backRes = MATCHES[(size_t)numPairs - 1] + LZMA_NUM_REPS; MOVE_POS(p, mainLen - 1) return mainLen; } @@ -1298,13 +1395,13 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position) if (len < 2) len = 2; else - while (len > matches[offs]) + while (len > MATCHES[offs]) offs += 2; for (; ; len++) { COptimal *opt; - UInt32 dist = matches[(size_t)offs + 1]; + UInt32 dist = MATCHES[(size_t)offs + 1]; UInt32 price = normalMatchPrice + GET_PRICE_LEN(&p->lenEnc, posState, len); unsigned lenToPosState = GetLenToPosState(len); @@ -1328,7 +1425,7 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position) opt->extra = 0; } - if (len == matches[offs]) + if (len == MATCHES[offs]) { offs += 2; if (offs == numPairs) @@ -1749,8 +1846,8 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position) if (newLen > numAvail) { newLen = numAvail; - for (numPairs = 0; newLen > matches[numPairs]; numPairs += 2); - matches[numPairs] = (UInt32)newLen; + for (numPairs = 0; newLen > MATCHES[numPairs]; numPairs += 2); + MATCHES[numPairs] = (UInt32)newLen; numPairs += 2; } @@ -1769,9 +1866,9 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position) } offs = 0; - while (startLen > matches[offs]) + while (startLen > MATCHES[offs]) offs += 2; - dist = matches[(size_t)offs + 1]; + dist = MATCHES[(size_t)offs + 1]; // if (dist >= kNumFullDistances) GetPosSlot2(dist, posSlot); @@ -1798,7 +1895,7 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position) } } - if (len == matches[offs]) + if (len == MATCHES[offs]) { // if (p->_maxMode) { // MATCH : LIT : REP_0 @@ -1863,7 +1960,7 @@ static unsigned GetOptimum(CLzmaEnc *p, UInt32 position) offs += 2; if (offs == numPairs) break; - dist = matches[(size_t)offs + 1]; + dist = MATCHES[(size_t)offs + 1]; // if (dist >= kNumFullDistances) GetPosSlot2(dist, posSlot); } @@ -2081,8 +2178,23 @@ static SRes CheckErrors(CLzmaEnc *p) return p->result; if (p->rc.res != SZ_OK) p->result = SZ_ERROR_WRITE; - if (p->matchFinderBase.result != SZ_OK) + + #ifndef _7ZIP_ST + if ( + // p->mf_Failure || + (p->mtMode && + ( // p->matchFinderMt.failure_LZ_LZ || + p->matchFinderMt.failure_LZ_BT)) + ) + { + p->result = MY_HRES_ERROR__INTERNAL_ERROR; + // printf("\nCheckErrors p->matchFinderMt.failureLZ\n"); + } + #endif + + if (MFB.result != SZ_OK) p->result = SZ_ERROR_READ; + if (p->result != SZ_OK) p->finished = True; return p->result; @@ -2223,11 +2335,11 @@ MY_NO_INLINE static void FillDistancesPrices(CLzmaEnc *p) static void LzmaEnc_Construct(CLzmaEnc *p) { RangeEnc_Construct(&p->rc); - MatchFinder_Construct(&p->matchFinderBase); + MatchFinder_Construct(&MFB); #ifndef _7ZIP_ST + p->matchFinderMt.MatchFinder = &MFB; MatchFinderMt_Construct(&p->matchFinderMt); - p->matchFinderMt.MatchFinder = &p->matchFinderBase; #endif { @@ -2243,7 +2355,6 @@ static void LzmaEnc_Construct(CLzmaEnc *p) LzmaEnc_InitPriceTables(p->ProbPrices); p->litProbs = NULL; p->saveState.litProbs = NULL; - } CLzmaEncHandle LzmaEnc_Create(ISzAllocPtr alloc) @@ -2269,7 +2380,7 @@ static void LzmaEnc_Destruct(CLzmaEnc *p, ISzAllocPtr alloc, ISzAllocPtr allocBi MatchFinderMt_Destruct(&p->matchFinderMt, allocBig); #endif - MatchFinder_Free(&p->matchFinderBase, allocBig); + MatchFinder_Free(&MFB, allocBig); LzmaEnc_FreeLits(p, alloc); RangeEnc_Free(&p->rc, alloc); } @@ -2287,6 +2398,12 @@ static SRes LzmaEnc_CodeOneBlock(CLzmaEnc *p, UInt32 maxPackSize, UInt32 maxUnpa UInt32 nowPos32, startPos32; if (p->needInit) { + #ifndef _7ZIP_ST + if (p->mtMode) + { + RINOK(MatchFinderMt_InitMt(&p->matchFinderMt)); + } + #endif p->matchFinder.Init(p->matchFinderObj); p->needInit = 0; } @@ -2582,11 +2699,13 @@ static SRes LzmaEnc_CodeOneBlock(CLzmaEnc *p, UInt32 maxPackSize, UInt32 maxUnpa static SRes LzmaEnc_Alloc(CLzmaEnc *p, UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig) { UInt32 beforeSize = kNumOpts; + UInt32 dictSize; + if (!RangeEnc_Alloc(&p->rc, alloc)) return SZ_ERROR_MEM; #ifndef _7ZIP_ST - p->mtMode = (p->multiThread && !p->fastMode && (p->matchFinderBase.btMode != 0)); + p->mtMode = (p->multiThread && !p->fastMode && (MFB.btMode != 0)); #endif { @@ -2605,30 +2724,50 @@ static SRes LzmaEnc_Alloc(CLzmaEnc *p, UInt32 keepWindowSize, ISzAllocPtr alloc, } } - p->matchFinderBase.bigHash = (Byte)(p->dictSize > kBigHashDicLimit ? 1 : 0); + MFB.bigHash = (Byte)(p->dictSize > kBigHashDicLimit ? 1 : 0); - if (beforeSize + p->dictSize < keepWindowSize) - beforeSize = keepWindowSize - p->dictSize; + + dictSize = p->dictSize; + if (dictSize == ((UInt32)2 << 30) || + dictSize == ((UInt32)3 << 30)) + { + /* 21.03 : here we reduce the dictionary for 2 reasons: + 1) we don't want 32-bit back_distance matches in decoder for 2 GB dictionary. + 2) we want to elimate useless last MatchFinder_Normalize3() for corner cases, + where data size is aligned for 1 GB: 5/6/8 GB. + That reducing must be >= 1 for such corner cases. */ + dictSize -= 1; + } + + if (beforeSize + dictSize < keepWindowSize) + beforeSize = keepWindowSize - dictSize; + + /* in worst case we can look ahead for + max(LZMA_MATCH_LEN_MAX, numFastBytes + 1 + numFastBytes) bytes. + we send larger value for (keepAfter) to MantchFinder_Create(): + (numFastBytes + LZMA_MATCH_LEN_MAX + 1) + */ #ifndef _7ZIP_ST if (p->mtMode) { - RINOK(MatchFinderMt_Create(&p->matchFinderMt, p->dictSize, beforeSize, p->numFastBytes, - LZMA_MATCH_LEN_MAX - + 1 /* 18.04 */ + RINOK(MatchFinderMt_Create(&p->matchFinderMt, dictSize, beforeSize, + p->numFastBytes, LZMA_MATCH_LEN_MAX + 1 /* 18.04 */ , allocBig)); p->matchFinderObj = &p->matchFinderMt; - p->matchFinderBase.bigHash = (Byte)( - (p->dictSize > kBigHashDicLimit && p->matchFinderBase.hashMask >= 0xFFFFFF) ? 1 : 0); + MFB.bigHash = (Byte)( + (p->dictSize > kBigHashDicLimit && MFB.hashMask >= 0xFFFFFF) ? 1 : 0); MatchFinderMt_CreateVTable(&p->matchFinderMt, &p->matchFinder); } else #endif { - if (!MatchFinder_Create(&p->matchFinderBase, p->dictSize, beforeSize, p->numFastBytes, LZMA_MATCH_LEN_MAX, allocBig)) + if (!MatchFinder_Create(&MFB, dictSize, beforeSize, + p->numFastBytes, LZMA_MATCH_LEN_MAX + 1 /* 21.03 */ + , allocBig)) return SZ_ERROR_MEM; - p->matchFinderObj = &p->matchFinderBase; - MatchFinder_CreateVTable(&p->matchFinderBase, &p->matchFinder); + p->matchFinderObj = &MFB; + MatchFinder_CreateVTable(&MFB, &p->matchFinder); } return SZ_OK; @@ -2700,6 +2839,8 @@ static void LzmaEnc_Init(CLzmaEnc *p) p->pbMask = ((unsigned)1 << p->pb) - 1; p->lpMask = ((UInt32)0x100 << p->lp) - ((unsigned)0x100 >> p->lc); + + // p->mf_Failure = False; } @@ -2742,7 +2883,7 @@ static SRes LzmaEnc_Prepare(CLzmaEncHandle pp, ISeqOutStream *outStream, ISeqInS ISzAllocPtr alloc, ISzAllocPtr allocBig) { CLzmaEnc *p = (CLzmaEnc *)pp; - p->matchFinderBase.stream = inStream; + MFB.stream = inStream; p->needInit = 1; p->rc.outStream = outStream; return LzmaEnc_AllocAndInit(p, 0, alloc, allocBig); @@ -2753,16 +2894,16 @@ SRes LzmaEnc_PrepareForLzma2(CLzmaEncHandle pp, ISzAllocPtr alloc, ISzAllocPtr allocBig) { CLzmaEnc *p = (CLzmaEnc *)pp; - p->matchFinderBase.stream = inStream; + MFB.stream = inStream; p->needInit = 1; return LzmaEnc_AllocAndInit(p, keepWindowSize, alloc, allocBig); } static void LzmaEnc_SetInputBuf(CLzmaEnc *p, const Byte *src, SizeT srcLen) { - p->matchFinderBase.directInput = 1; - p->matchFinderBase.bufferBase = (Byte *)src; - p->matchFinderBase.directInputRem = srcLen; + MFB.directInput = 1; + MFB.bufferBase = (Byte *)src; + MFB.directInputRem = srcLen; } SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen, @@ -2804,9 +2945,12 @@ static size_t SeqOutStreamBuf_Write(const ISeqOutStream *pp, const void *data, s size = p->rem; p->overflow = True; } - memcpy(p->data, data, size); - p->rem -= size; - p->data += size; + if (size != 0) + { + memcpy(p->data, data, size); + p->rem -= size; + p->data += size; + } return size; } @@ -2826,6 +2970,7 @@ const Byte *LzmaEnc_GetCurBuf(CLzmaEncHandle pp) } +// (desiredPackSize == 0) is not allowed SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit, Byte *dest, size_t *destLen, UInt32 desiredPackSize, UInt32 *unpackSize) { @@ -2846,14 +2991,10 @@ SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit, if (reInit) LzmaEnc_Init(p); LzmaEnc_InitPrices(p); - - nowPos64 = p->nowPos64; RangeEnc_Init(&p->rc); p->rc.outStream = &outStream.vt; - - if (desiredPackSize == 0) - return SZ_ERROR_OUTPUT_EOF; - + nowPos64 = p->nowPos64; + res = LzmaEnc_CodeOneBlock(p, desiredPackSize, *unpackSize); *unpackSize = (UInt32)(p->nowPos64 - nowPos64); @@ -2895,7 +3036,7 @@ static SRes LzmaEnc_Encode2(CLzmaEnc *p, ICompressProgress *progress) LzmaEnc_Finish(p); /* - if (res == SZ_OK && !Inline_MatchFinder_IsFinishedOK(&p->matchFinderBase)) + if (res == SZ_OK && !Inline_MatchFinder_IsFinishedOK(&MFB)) res = SZ_ERROR_FAIL; } */ @@ -2914,29 +3055,37 @@ SRes LzmaEnc_Encode(CLzmaEncHandle pp, ISeqOutStream *outStream, ISeqInStream *i SRes LzmaEnc_WriteProperties(CLzmaEncHandle pp, Byte *props, SizeT *size) { - CLzmaEnc *p = (CLzmaEnc *)pp; - unsigned i; - UInt32 dictSize = p->dictSize; if (*size < LZMA_PROPS_SIZE) return SZ_ERROR_PARAM; *size = LZMA_PROPS_SIZE; - props[0] = (Byte)((p->pb * 5 + p->lp) * 9 + p->lc); - - if (dictSize >= ((UInt32)1 << 22)) { - const UInt32 kDictMask = ((UInt32)1 << 20) - 1; - if (dictSize < (UInt32)0xFFFFFFFF - kDictMask) - dictSize = (dictSize + kDictMask) & ~kDictMask; - } - else for (i = 11; i <= 30; i++) - { - if (dictSize <= ((UInt32)2 << i)) { dictSize = ((UInt32)2 << i); break; } - if (dictSize <= ((UInt32)3 << i)) { dictSize = ((UInt32)3 << i); break; } - } + const CLzmaEnc *p = (const CLzmaEnc *)pp; + const UInt32 dictSize = p->dictSize; + UInt32 v; + props[0] = (Byte)((p->pb * 5 + p->lp) * 9 + p->lc); + + // we write aligned dictionary value to properties for lzma decoder + if (dictSize >= ((UInt32)1 << 21)) + { + const UInt32 kDictMask = ((UInt32)1 << 20) - 1; + v = (dictSize + kDictMask) & ~kDictMask; + if (v < dictSize) + v = dictSize; + } + else + { + unsigned i = 11 * 2; + do + { + v = (UInt32)(2 + (i & 1)) << (i >> 1); + i++; + } + while (v < dictSize); + } - for (i = 0; i < 4; i++) - props[1 + i] = (Byte)(dictSize >> (8 * i)); - return SZ_OK; + SetUi32(props + 1, v); + return SZ_OK; + } } diff --git a/deps/LZMA-SDK/C/MtCoder.c b/deps/LZMA-SDK/C/MtCoder.c index 85444f484..e39d9cb19 100644 --- a/deps/LZMA-SDK/C/MtCoder.c +++ b/deps/LZMA-SDK/C/MtCoder.c @@ -1,5 +1,5 @@ /* MtCoder.c -- Multi-thread Coder -2021-02-09 : Igor Pavlov : Public domain */ +2021-12-21 : Igor Pavlov : Public domain */ #include "Precomp.h" @@ -44,7 +44,7 @@ static WRes ArEvent_OptCreate_And_Reset(CEvent *p) } -static THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE ThreadFunc(void *pp); +static THREAD_FUNC_DECL ThreadFunc(void *pp); static SRes MtCoderThread_CreateAndStart(CMtCoderThread *t) @@ -335,7 +335,7 @@ static SRes ThreadFunc2(CMtCoderThread *t) } -static THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE ThreadFunc(void *pp) +static THREAD_FUNC_DECL ThreadFunc(void *pp) { CMtCoderThread *t = (CMtCoderThread *)pp; for (;;) @@ -495,12 +495,7 @@ SRes MtCoder_Code(CMtCoder *p) { RINOK_THREAD(ArEvent_OptCreate_And_Reset(&p->readEvent)); - - if (Semaphore_IsCreated(&p->blocksSemaphore)) - { - RINOK_THREAD(Semaphore_Close(&p->blocksSemaphore)); - } - RINOK_THREAD(Semaphore_Create(&p->blocksSemaphore, numBlocksMax, numBlocksMax)); + RINOK_THREAD(Semaphore_OptCreateInit(&p->blocksSemaphore, numBlocksMax, numBlocksMax)); } for (i = 0; i < MTCODER__BLOCKS_MAX - 1; i++) diff --git a/deps/LZMA-SDK/C/MtDec.c b/deps/LZMA-SDK/C/MtDec.c index 24441b3a7..854087b92 100644 --- a/deps/LZMA-SDK/C/MtDec.c +++ b/deps/LZMA-SDK/C/MtDec.c @@ -1,5 +1,5 @@ /* MtDec.c -- Multi-thread Decoder -2021-02-27 : Igor Pavlov : Public domain */ +2021-12-21 : Igor Pavlov : Public domain */ #include "Precomp.h" @@ -102,7 +102,7 @@ typedef struct __CMtDecBufLink CMtDecBufLink; -static THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE ThreadFunc(void *pp); +static THREAD_FUNC_DECL ThreadFunc(void *pp); static WRes MtDecThread_CreateEvents(CMtDecThread *t) @@ -836,7 +836,7 @@ static WRes ThreadFunc2(CMtDecThread *t) #endif -static THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE ThreadFunc1(void *pp) +static THREAD_FUNC_DECL ThreadFunc1(void *pp) { WRes res; @@ -862,7 +862,7 @@ static THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE ThreadFunc1(void *pp) return (THREAD_FUNC_RET_TYPE)(UINT_PTR)res; } -static MY_NO_INLINE THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE ThreadFunc(void *pp) +static MY_NO_INLINE THREAD_FUNC_DECL ThreadFunc(void *pp) { #ifdef USE_ALLOCA CMtDecThread *t = (CMtDecThread *)pp; diff --git a/deps/LZMA-SDK/C/Threads.c b/deps/LZMA-SDK/C/Threads.c index 402abab01..6eb45b08a 100644 --- a/deps/LZMA-SDK/C/Threads.c +++ b/deps/LZMA-SDK/C/Threads.c @@ -1,11 +1,11 @@ /* Threads.c -- multithreading library -2021-04-25 : Igor Pavlov : Public domain */ +2021-12-21 : Igor Pavlov : Public domain */ #include "Precomp.h" #ifdef _WIN32 -#ifndef UNDER_CE +#ifndef USE_THREADS_CreateThread #include #endif @@ -63,10 +63,10 @@ WRes Thread_Create(CThread *p, THREAD_FUNC_TYPE func, LPVOID param) { /* Windows Me/98/95: threadId parameter may not be NULL in _beginthreadex/CreateThread functions */ - #ifdef UNDER_CE + #ifdef USE_THREADS_CreateThread DWORD threadId; - *p = CreateThread(0, 0, func, param, 0, &threadId); + *p = CreateThread(NULL, 0, func, param, 0, &threadId); #else @@ -82,7 +82,7 @@ WRes Thread_Create(CThread *p, THREAD_FUNC_TYPE func, LPVOID param) WRes Thread_Create_With_Affinity(CThread *p, THREAD_FUNC_TYPE func, LPVOID param, CAffinityMask affinity) { - #ifdef UNDER_CE + #ifdef USE_THREADS_CreateThread UNUSED_VAR(affinity) return Thread_Create(p, func, param); @@ -150,6 +150,17 @@ WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount) return HandleToWRes(*p); } +WRes Semaphore_OptCreateInit(CSemaphore *p, UInt32 initCount, UInt32 maxCount) +{ + // if (Semaphore_IsCreated(p)) + { + WRes wres = Semaphore_Close(p); + if (wres != 0) + return wres; + } + return Semaphore_Create(p, initCount, maxCount); +} + static WRes Semaphore_Release(CSemaphore *p, LONG releaseCount, LONG *previousCount) { return BOOLToWRes(ReleaseSemaphore(*p, releaseCount, previousCount)); } WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 num) @@ -158,7 +169,9 @@ WRes Semaphore_Release1(CSemaphore *p) { return Semaphore_ReleaseN(p, 1); } WRes CriticalSection_Init(CCriticalSection *p) { - /* InitializeCriticalSection can raise only STATUS_NO_MEMORY exception */ + /* InitializeCriticalSection() can raise exception: + Windows XP, 2003 : can raise a STATUS_NO_MEMORY exception + Windows Vista+ : no exceptions */ #ifdef _MSC_VER __try #endif @@ -167,7 +180,7 @@ WRes CriticalSection_Init(CCriticalSection *p) /* InitializeCriticalSectionAndSpinCount(p, 0); */ } #ifdef _MSC_VER - __except (EXCEPTION_EXECUTE_HANDLER) { return 1; } + __except (EXCEPTION_EXECUTE_HANDLER) { return ERROR_NOT_ENOUGH_MEMORY; } #endif return 0; } @@ -406,6 +419,27 @@ WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount) return 0; } + +WRes Semaphore_OptCreateInit(CSemaphore *p, UInt32 initCount, UInt32 maxCount) +{ + if (Semaphore_IsCreated(p)) + { + /* + WRes wres = Semaphore_Close(p); + if (wres != 0) + return wres; + */ + if (initCount > maxCount || maxCount < 1) + return EINVAL; + // return EINVAL; // for debug + p->_count = initCount; + p->_maxCount = maxCount; + return 0; + } + return Semaphore_Create(p, initCount, maxCount); +} + + WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 releaseCount) { UInt32 newCount; diff --git a/deps/LZMA-SDK/C/Threads.h b/deps/LZMA-SDK/C/Threads.h index c555c8b97..e9493afff 100644 --- a/deps/LZMA-SDK/C/Threads.h +++ b/deps/LZMA-SDK/C/Threads.h @@ -1,21 +1,25 @@ /* Threads.h -- multithreading library -2021-04-25 : Igor Pavlov : Public domain */ +2021-12-21 : Igor Pavlov : Public domain */ #ifndef __7Z_THREADS_H #define __7Z_THREADS_H #ifdef _WIN32 -#include +#include #else -#if !defined(__APPLE__) && !defined(_AIX) +#if defined(__linux__) +#if !defined(__APPLE__) && !defined(_AIX) && !defined(__ANDROID__) #ifndef _7ZIP_AFFINITY_DISABLE #define _7ZIP_AFFINITY_SUPPORTED +// #pragma message(" ==== _7ZIP_AFFINITY_SUPPORTED") // #define _GNU_SOURCE #endif #endif +#endif #include + #endif #include "7zTypes.h" @@ -34,8 +38,14 @@ typedef HANDLE CThread; #define Thread_Close(p) HandlePtr_Close(p) // #define Thread_Wait(p) Handle_WaitObject(*(p)) +#ifdef UNDER_CE + // if (USE_THREADS_CreateThread is defined), we use _beginthreadex() + // if (USE_THREADS_CreateThread is not definned), we use CreateThread() + #define USE_THREADS_CreateThread +#endif + typedef - #ifdef UNDER_CE + #ifdef USE_THREADS_CreateThread DWORD #else unsigned @@ -86,7 +96,30 @@ typedef UInt64 CCpuSet; #define THREAD_FUNC_CALL_TYPE MY_STD_CALL -#define THREAD_FUNC_DECL THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE + +#if defined(_WIN32) && defined(__GNUC__) +/* GCC compiler for x86 32-bit uses the rule: + the stack is 16-byte aligned before CALL instruction for function calling. + But only root function main() contains instructions that + set 16-byte alignment for stack pointer. And another functions + just keep alignment, if it was set in some parent function. + + The problem: + if we create new thread in MinGW (GCC) 32-bit x86 via _beginthreadex() or CreateThread(), + the root function of thread doesn't set 16-byte alignment. + And stack frames in all child functions also will be unaligned in that case. + + Here we set (force_align_arg_pointer) attribute for root function of new thread. + Do we need (force_align_arg_pointer) also for another systems? */ + + #define THREAD_FUNC_ATTRIB_ALIGN_ARG __attribute__((force_align_arg_pointer)) + // #define THREAD_FUNC_ATTRIB_ALIGN_ARG // for debug : bad alignment in SSE functions +#else + #define THREAD_FUNC_ATTRIB_ALIGN_ARG +#endif + +#define THREAD_FUNC_DECL THREAD_FUNC_ATTRIB_ALIGN_ARG THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE + typedef THREAD_FUNC_RET_TYPE (THREAD_FUNC_CALL_TYPE * THREAD_FUNC_TYPE)(void *); WRes Thread_Create(CThread *p, THREAD_FUNC_TYPE func, LPVOID param); WRes Thread_Create_With_Affinity(CThread *p, THREAD_FUNC_TYPE func, LPVOID param, CAffinityMask affinity); @@ -122,6 +155,7 @@ typedef HANDLE CSemaphore; #define Semaphore_Close(p) HandlePtr_Close(p) #define Semaphore_Wait(p) Handle_WaitObject(*(p)) WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount); +WRes Semaphore_OptCreateInit(CSemaphore *p, UInt32 initCount, UInt32 maxCount); WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 num); WRes Semaphore_Release1(CSemaphore *p); @@ -172,6 +206,7 @@ typedef struct _CSemaphore #define Semaphore_IsCreated(p) ((p)->_created) WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount); +WRes Semaphore_OptCreateInit(CSemaphore *p, UInt32 initCount, UInt32 maxCount); WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 num); #define Semaphore_Release1(p) Semaphore_ReleaseN(p, 1) WRes Semaphore_Wait(CSemaphore *p); diff --git a/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.c b/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.c index 6b4293e33..4f05f1e73 100644 --- a/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.c +++ b/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.c @@ -1,5 +1,5 @@ /* LzmaUtil.c -- Test application for LZMA compression -2021-02-15 : Igor Pavlov : Public domain */ +2021-11-01 : Igor Pavlov : Public domain */ #include "../../Precomp.h" @@ -12,6 +12,7 @@ #include "../../Alloc.h" #include "../../7zFile.h" #include "../../7zVersion.h" +#include "../../LzFind.h" #include "../../LzmaDec.h" #include "../../LzmaEnc.h" @@ -195,6 +196,8 @@ static int main2(int numArgs, const char *args[], char *rs) int encodeMode; BoolInt useOutFile = False; + LzFindPrepare(); + FileSeqInStream_CreateVTable(&inStream); File_Construct(&inStream.file); inStream.wres = 0; @@ -276,7 +279,7 @@ static int main2(int numArgs, const char *args[], char *rs) int MY_CDECL main(int numArgs, const char *args[]) { - char rs[800] = { 0 }; + char rs[1000] = { 0 }; int res = main2(numArgs, args, rs); fputs(rs, stdout); return res; diff --git a/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsp b/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsp index eedde07d8..44835d598 100644 --- a/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsp +++ b/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsp @@ -134,6 +134,10 @@ SOURCE=..\..\LzFindMt.h # End Source File # Begin Source File +SOURCE=..\..\LzFindOpt.c +# End Source File +# Begin Source File + SOURCE=..\..\LzHash.h # End Source File # Begin Source File diff --git a/deps/LZMA-SDK/C/Util/Lzma/makefile b/deps/LZMA-SDK/C/Util/Lzma/makefile index 3b825f21a..9d971f666 100644 --- a/deps/LZMA-SDK/C/Util/Lzma/makefile +++ b/deps/LZMA-SDK/C/Util/Lzma/makefile @@ -8,8 +8,10 @@ LIB_OBJS = \ C_OBJS = \ $O\Alloc.obj \ + $O\CpuArch.obj \ $O\LzFind.obj \ $O\LzFindMt.obj \ + $O\LzFindOpt.obj \ $O\LzmaDec.obj \ $O\LzmaEnc.obj \ $O\7zFile.obj \ diff --git a/deps/LZMA-SDK/C/Util/Lzma/makefile.gcc b/deps/LZMA-SDK/C/Util/Lzma/makefile.gcc index 89b3e11f7..6ce77a0f2 100644 --- a/deps/LZMA-SDK/C/Util/Lzma/makefile.gcc +++ b/deps/LZMA-SDK/C/Util/Lzma/makefile.gcc @@ -8,8 +8,10 @@ OBJS = \ $O/7zFile.o \ $O/7zStream.o \ $O/Alloc.o \ + $O/CpuArch.o \ $O/LzFind.o \ $O/LzFindMt.o \ + $O/LzFindOpt.o \ $O/LzmaDec.o \ $O/LzmaEnc.o \ $O/LzmaUtil.o \ diff --git a/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsp b/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsp index 0d4c981c4..c267bbef6 100644 --- a/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsp +++ b/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsp @@ -136,6 +136,10 @@ SOURCE=..\..\LzFindMt.h # End Source File # Begin Source File +SOURCE=..\..\LzFindOpt.c +# End Source File +# Begin Source File + SOURCE=..\..\LzHash.h # End Source File # Begin Source File diff --git a/deps/LZMA-SDK/C/Util/LzmaLib/makefile b/deps/LZMA-SDK/C/Util/LzmaLib/makefile index e0f311471..bcb7496be 100644 --- a/deps/LZMA-SDK/C/Util/LzmaLib/makefile +++ b/deps/LZMA-SDK/C/Util/LzmaLib/makefile @@ -11,8 +11,10 @@ LIB_OBJS = \ C_OBJS = \ $O\Alloc.obj \ + $O\CpuArch.obj \ $O\LzFind.obj \ $O\LzFindMt.obj \ + $O\LzFindOpt.obj \ $O\LzmaDec.obj \ $O\LzmaEnc.obj \ $O\LzmaLib.obj \ diff --git a/deps/LZMA-SDK/C/XzDec.c b/deps/LZMA-SDK/C/XzDec.c index d345f68c1..49329f16a 100644 --- a/deps/LZMA-SDK/C/XzDec.c +++ b/deps/LZMA-SDK/C/XzDec.c @@ -1,5 +1,5 @@ /* XzDec.c -- Xz Decode -2021-04-01 : Igor Pavlov : Public domain */ +2021-09-04 : Igor Pavlov : Public domain */ #include "Precomp.h" @@ -773,7 +773,8 @@ static BoolInt Xz_CheckFooter(CXzStreamFlags flags, UInt64 indexSize, const Byte #define READ_VARINT_AND_CHECK(buf, pos, size, res) \ { unsigned s = Xz_ReadVarInt(buf + pos, size - pos, res); \ - if (s == 0) return SZ_ERROR_ARCHIVE; pos += s; } + if (s == 0) return SZ_ERROR_ARCHIVE; \ + pos += s; } static BoolInt XzBlock_AreSupportedFilters(const CXzBlock *p) diff --git a/deps/LZMA-SDK/C/XzIn.c b/deps/LZMA-SDK/C/XzIn.c index 54d81c4a4..07201b842 100644 --- a/deps/LZMA-SDK/C/XzIn.c +++ b/deps/LZMA-SDK/C/XzIn.c @@ -1,5 +1,5 @@ /* XzIn.c - Xz input -2021-04-01 : Igor Pavlov : Public domain */ +2021-09-04 : Igor Pavlov : Public domain */ #include "Precomp.h" @@ -26,7 +26,8 @@ SRes Xz_ReadHeader(CXzStreamFlags *p, ISeqInStream *inStream) #define READ_VARINT_AND_CHECK(buf, pos, size, res) \ { unsigned s = Xz_ReadVarInt(buf + pos, size - pos, res); \ - if (s == 0) return SZ_ERROR_ARCHIVE; pos += s; } + if (s == 0) return SZ_ERROR_ARCHIVE; \ + pos += s; } SRes XzBlock_ReadHeader(CXzBlock *p, ISeqInStream *inStream, BoolInt *isIndex, UInt32 *headerSizeRes) { diff --git a/deps/LZMA-SDK/C/var_clang_x64.mak b/deps/LZMA-SDK/C/var_clang_x64.mak index d9013e1cd..b35f0cca4 100644 --- a/deps/LZMA-SDK/C/var_clang_x64.mak +++ b/deps/LZMA-SDK/C/var_clang_x64.mak @@ -9,4 +9,3 @@ USE_ASM=1 CC=$(CROSS_COMPILE)clang CXX=$(CROSS_COMPILE)clang++ USE_CLANG=1 - diff --git a/deps/LZMA-SDK/C/var_clang_x86.mak b/deps/LZMA-SDK/C/var_clang_x86.mak index 9ab916a70..d18f0ae12 100644 --- a/deps/LZMA-SDK/C/var_clang_x86.mak +++ b/deps/LZMA-SDK/C/var_clang_x86.mak @@ -9,4 +9,3 @@ USE_ASM=1 CC=$(CROSS_COMPILE)clang CXX=$(CROSS_COMPILE)clang++ USE_CLANG=1 - diff --git a/deps/LZMA-SDK/C/var_gcc_x86.mak b/deps/LZMA-SDK/C/var_gcc_x86.mak index 9eada64e1..3adf4a84a 100644 --- a/deps/LZMA-SDK/C/var_gcc_x86.mak +++ b/deps/LZMA-SDK/C/var_gcc_x86.mak @@ -8,4 +8,3 @@ MY_ARCH=-m32 USE_ASM=1 CC=$(CROSS_COMPILE)gcc CXX=$(CROSS_COMPILE)g++ - diff --git a/deps/LZMA-SDK/C/warn_gcc.mak b/deps/LZMA-SDK/C/warn_gcc.mak index 3ae796480..8fbdb28f6 100644 --- a/deps/LZMA-SDK/C/warn_gcc.mak +++ b/deps/LZMA-SDK/C/warn_gcc.mak @@ -49,5 +49,3 @@ CFLAGS_WARN_GCC_PPMD_UNALIGNED = \ CFLAGS_WARN = $(CFLAGS_WARN_GCC_9) \ # $(CFLAGS_WARN_GCC_PPMD_UNALIGNED) - - \ No newline at end of file diff --git a/deps/LZMA-SDK/DOC/lzma-history.txt b/deps/LZMA-SDK/DOC/lzma-history.txt index 3fc19fd8b..9c5a2416c 100644 --- a/deps/LZMA-SDK/DOC/lzma-history.txt +++ b/deps/LZMA-SDK/DOC/lzma-history.txt @@ -1,6 +1,42 @@ HISTORY of the LZMA SDK ----------------------- +21.07 2021-12-26 +------------------------- +- New switches: -spm and -im!{file_path} to exclude directories from processing + for specified paths that don't contain path separator character at the end of path. +- The sorting order of files in archives was slightly changed to be more consistent + for cases where the name of some directory is the same as the prefix part of the name + of another directory or file. + + +21.06 2021-11-24 +------------------------- +- Bug in LZMA encoder in file LzmaEnc.c was fixed: + LzmaEnc_MemEncode(), LzmaEncode() and LzmaCompress() could work incorrectly, + if size value for output buffer is smaller than size required for all compressed data. + LzmaEnc_Encode() could work incorrectly, + if callback ISeqOutStream::Write() doesn't write all compressed data. + NCompress::NLzma::CEncoder::Code() could work incorrectly, + if callback ISequentialOutStream::Write() returns error code. +- Bug in versions 21.00-21.05 was fixed: + 7-Zip didn't set attributes of directories during archive extracting. + + +21.04 beta 2021-11-02 +------------------------- +- 7-Zip now reduces the number of working CPU threads for compression, + if RAM size is not enough for compression with big LZMA2 dictionary. +- 7-Zip now can create and check "file.sha256" text files that contain the list + of file names and SHA-256 checksums in format compatible with sha256sum program. + + +21.03 beta 2021-07-20 +------------------------- +- The maximum dictionary size for LZMA/LZMA2 compressing was increased to 4 GB (3840 MiB). +- Minor speed optimizations in LZMA/LZMA2 compressing. + + 21.02 alpha 2021-05-06 ------------------------- - The command line version of 7-Zip for macOS was released. diff --git a/deps/LZMA-SDK/DOC/lzma-sdk.txt b/deps/LZMA-SDK/DOC/lzma-sdk.txt index b11716938..d54aad328 100644 --- a/deps/LZMA-SDK/DOC/lzma-sdk.txt +++ b/deps/LZMA-SDK/DOC/lzma-sdk.txt @@ -1,4 +1,4 @@ -LZMA SDK 21.02 +LZMA SDK 22.01 -------------- LZMA SDK provides the documentation, samples, header files, @@ -62,14 +62,61 @@ LZMA SDK Contents UNIX/Linux version ------------------ -To compile C++ version of file->file LZMA encoding, go to directory -CPP/7zip/Bundles/LzmaCon -and call make to recompile it: - make -f makefile.gcc clean all +There are several otpions to compile 7-Zip with different compilers: gcc and clang. +Also 7-Zip code contains two versions for some critical parts of code: in C and in Assembeler. +So if you compile the version with Assembeler code, you will get faster 7-Zip binary. + +7-Zip's assembler code uses the following syntax for different platforms: + +1) x86 and x86-64 (AMD64): MASM syntax. + There are 2 programs that supports MASM syntax in Linux. +' 'Asmc Macro Assembler and JWasm. But JWasm now doesn't support some + cpu instructions used in 7-Zip. + So you must install Asmc Macro Assembler in Linux, if you want to compile fastest version + of 7-Zip x86 and x86-64: + https://github.com/nidud/asmc + +2) arm64: GNU assembler for ARM64 with preprocessor. + That systax of that arm64 assembler code in 7-Zip is supported by GCC and CLANG for ARM64. + +There are different binaries that can be compiled from 7-Zip source. +There are 2 main files in folder for compiling: + makefile - that can be used for compiling Windows version of 7-Zip with nmake command + makefile.gcc - that can be used for compiling Linux/macOS versions of 7-Zip with make command + +At first you must change the current folder to folder that contains `makefile.gcc`: + + cd CPP/7zip/Bundles/Alone7z + +Then you can compile `makefile.gcc` with the command: + + make -j -f makefile.gcc + +Also there are additional "*.mak" files in folder "CPP/7zip/" that can be used to compile +7-Zip binaries with optimized code and optimzing options. + +To compile with GCC without assembler: + cd CPP/7zip/Bundles/Alone7z + make -j -f ../../cmpl_gcc.mak + +To compile with CLANG without assembler: + make -j -f ../../cmpl_clang.mak + +To compile 7-Zip for x86-64 with asmc assembler: + make -j -f ../../cmpl_gcc_x64.mak + +To compile 7-Zip for arm64 with assembler: + make -j -f ../../cmpl_gcc_arm64.mak + +To compile 7-Zip for arm64 for macOS: + make -j -f ../../cmpl_mac_arm64.mak + +Also you can change some compiler options in the mak files: + cmpl_gcc.mak + var_gcc.mak + warn_gcc.mak + -In some UNIX/Linux versions you must compile LZMA with static libraries. -To compile with static libraries, you can use -LIB = -lm -static Also you can use p7zip (port of 7-Zip for POSIX systems like Unix or Linux): diff --git a/docs/changes.txt b/docs/changes.txt index a61dbe080..78deba30e 100644 --- a/docs/changes.txt +++ b/docs/changes.txt @@ -77,6 +77,7 @@ - Apple Driver: Updated requirements to use Apple OpenCL API to macOS 13.0 - use - Backend Checks: Describe workaround in error message when detecting more than 64 backend devices - Brain: Added sanity check and corresponding error message for invalid --brain-port values +- Dependencies: Updated LZMA SDK to 22.01 - Modules: Added support for non-zero IVs for -m 6800 (Lastpass). Also added `tools/lastpass2hashcat.py` - Open Document Format: Added support for small documents with content length < 1024 - Status Code: Add specific return code for self-test fail (-11) From ecc03008a02686783f0404e5d44e3bd45693fbdc Mon Sep 17 00:00:00 2001 From: Gabriele Gristina Date: Wed, 17 May 2023 19:00:48 +0200 Subject: [PATCH 2/3] fix LZMA_SDK Windows.h includes --- deps/LZMA-SDK/C/Alloc.c | 2 +- deps/LZMA-SDK/C/CpuArch.c | 6 +++--- deps/LZMA-SDK/C/DllSecur.c | 2 +- deps/LZMA-SDK/C/Threads.h | 2 +- 4 files changed, 6 insertions(+), 6 deletions(-) diff --git a/deps/LZMA-SDK/C/Alloc.c b/deps/LZMA-SDK/C/Alloc.c index 142a1ea22..0d0107c56 100644 --- a/deps/LZMA-SDK/C/Alloc.c +++ b/deps/LZMA-SDK/C/Alloc.c @@ -6,7 +6,7 @@ #include #ifdef _WIN32 -#include +#include #endif #include diff --git a/deps/LZMA-SDK/C/CpuArch.c b/deps/LZMA-SDK/C/CpuArch.c index a0e93e8b0..36e0be0b1 100644 --- a/deps/LZMA-SDK/C/CpuArch.c +++ b/deps/LZMA-SDK/C/CpuArch.c @@ -217,7 +217,7 @@ BoolInt CPU_Is_InOrder() } #if !defined(MY_CPU_AMD64) && defined(_WIN32) -#include +#include static BoolInt CPU_Sys_Is_SSE_Supported() { OSVERSIONINFO vi; @@ -275,7 +275,7 @@ BoolInt CPU_IsSupported_SHA() // #include #ifdef _WIN32 -#include +#include #endif BoolInt CPU_IsSupported_AVX2() @@ -351,7 +351,7 @@ BoolInt CPU_IsSupported_PageGB() #ifdef _WIN32 -#include +#include BoolInt CPU_IsSupported_CRC32() { return IsProcessorFeaturePresent(PF_ARM_V8_CRC32_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } BoolInt CPU_IsSupported_CRYPTO() { return IsProcessorFeaturePresent(PF_ARM_V8_CRYPTO_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } diff --git a/deps/LZMA-SDK/C/DllSecur.c b/deps/LZMA-SDK/C/DllSecur.c index a37c1b3e2..16755bba9 100644 --- a/deps/LZMA-SDK/C/DllSecur.c +++ b/deps/LZMA-SDK/C/DllSecur.c @@ -5,7 +5,7 @@ #ifdef _WIN32 -#include +#include #include "DllSecur.h" diff --git a/deps/LZMA-SDK/C/Threads.h b/deps/LZMA-SDK/C/Threads.h index e9493afff..71972558d 100644 --- a/deps/LZMA-SDK/C/Threads.h +++ b/deps/LZMA-SDK/C/Threads.h @@ -5,7 +5,7 @@ #define __7Z_THREADS_H #ifdef _WIN32 -#include +#include #else #if defined(__linux__) From 468f91b166c1491d33944f2e13bb73a3019c5ab5 Mon Sep 17 00:00:00 2001 From: Gabriele Gristina Date: Wed, 17 May 2023 19:03:13 +0200 Subject: [PATCH 3/3] perform dos2unix for all LZMA SDK files --- deps/LZMA-SDK/C/7z.h | 408 +- deps/LZMA-SDK/C/7zAlloc.c | 160 +- deps/LZMA-SDK/C/7zAlloc.h | 38 +- deps/LZMA-SDK/C/7zArcIn.c | 3566 +++++----- deps/LZMA-SDK/C/7zBuf.c | 72 +- deps/LZMA-SDK/C/7zBuf.h | 70 +- deps/LZMA-SDK/C/7zBuf2.c | 104 +- deps/LZMA-SDK/C/7zCrc.c | 644 +- deps/LZMA-SDK/C/7zCrc.h | 50 +- deps/LZMA-SDK/C/7zCrcOpt.c | 234 +- deps/LZMA-SDK/C/7zDec.c | 1200 ++-- deps/LZMA-SDK/C/7zFile.c | 884 +-- deps/LZMA-SDK/C/7zFile.h | 182 +- deps/LZMA-SDK/C/7zStream.c | 352 +- deps/LZMA-SDK/C/7zTypes.h | 1058 +-- deps/LZMA-SDK/C/7zVersion.h | 54 +- deps/LZMA-SDK/C/7zVersion.rc | 110 +- deps/LZMA-SDK/C/7zip_gcc_c.mak | 686 +- deps/LZMA-SDK/C/Aes.c | 750 +- deps/LZMA-SDK/C/Aes.h | 120 +- deps/LZMA-SDK/C/AesOpt.c | 1552 ++-- deps/LZMA-SDK/C/Alloc.c | 926 +-- deps/LZMA-SDK/C/Alloc.h | 116 +- deps/LZMA-SDK/C/Bcj2.c | 514 +- deps/LZMA-SDK/C/Bcj2.h | 292 +- deps/LZMA-SDK/C/Bcj2Enc.c | 622 +- deps/LZMA-SDK/C/Bra.c | 460 +- deps/LZMA-SDK/C/Bra.h | 128 +- deps/LZMA-SDK/C/Bra86.c | 164 +- deps/LZMA-SDK/C/BraIA64.c | 106 +- deps/LZMA-SDK/C/Compiler.h | 86 +- deps/LZMA-SDK/C/CpuArch.c | 956 +-- deps/LZMA-SDK/C/CpuArch.h | 890 +-- deps/LZMA-SDK/C/Delta.c | 338 +- deps/LZMA-SDK/C/Delta.h | 38 +- deps/LZMA-SDK/C/DllSecur.c | 228 +- deps/LZMA-SDK/C/DllSecur.h | 40 +- deps/LZMA-SDK/C/LzFind.c | 3256 ++++----- deps/LZMA-SDK/C/LzFind.h | 272 +- deps/LZMA-SDK/C/LzFindMt.c | 2800 ++++---- deps/LZMA-SDK/C/LzFindMt.h | 218 +- deps/LZMA-SDK/C/LzFindOpt.c | 1156 +-- deps/LZMA-SDK/C/LzHash.h | 68 +- deps/LZMA-SDK/C/Lzma2Dec.c | 978 +-- deps/LZMA-SDK/C/Lzma2Dec.h | 240 +- deps/LZMA-SDK/C/Lzma2DecMt.c | 2180 +++--- deps/LZMA-SDK/C/Lzma2DecMt.h | 158 +- deps/LZMA-SDK/C/Lzma2Enc.c | 1606 ++--- deps/LZMA-SDK/C/Lzma2Enc.h | 110 +- deps/LZMA-SDK/C/Lzma86.h | 222 +- deps/LZMA-SDK/C/Lzma86Dec.c | 108 +- deps/LZMA-SDK/C/Lzma86Enc.c | 208 +- deps/LZMA-SDK/C/LzmaDec.c | 2726 +++---- deps/LZMA-SDK/C/LzmaDec.h | 472 +- deps/LZMA-SDK/C/LzmaEnc.c | 6324 ++++++++--------- deps/LZMA-SDK/C/LzmaEnc.h | 156 +- deps/LZMA-SDK/C/LzmaLib.c | 80 +- deps/LZMA-SDK/C/LzmaLib.h | 276 +- deps/LZMA-SDK/C/MtCoder.c | 1190 ++-- deps/LZMA-SDK/C/MtCoder.h | 282 +- deps/LZMA-SDK/C/MtDec.c | 2278 +++--- deps/LZMA-SDK/C/MtDec.h | 404 +- deps/LZMA-SDK/C/Ppmd.h | 334 +- deps/LZMA-SDK/C/Ppmd7.c | 2208 +++--- deps/LZMA-SDK/C/Ppmd7.h | 362 +- deps/LZMA-SDK/C/Ppmd7Dec.c | 594 +- deps/LZMA-SDK/C/Ppmd7Enc.c | 646 +- deps/LZMA-SDK/C/Precomp.h | 20 +- deps/LZMA-SDK/C/RotateDefs.h | 60 +- deps/LZMA-SDK/C/Sha256.c | 972 +-- deps/LZMA-SDK/C/Sha256.h | 152 +- deps/LZMA-SDK/C/Sha256Opt.c | 746 +- deps/LZMA-SDK/C/Sort.c | 282 +- deps/LZMA-SDK/C/Sort.h | 36 +- deps/LZMA-SDK/C/Threads.c | 1080 +-- deps/LZMA-SDK/C/Threads.h | 464 +- deps/LZMA-SDK/C/Util/7z/7z.dsp | 482 +- deps/LZMA-SDK/C/Util/7z/7z.dsw | 58 +- deps/LZMA-SDK/C/Util/7z/7zMain.c | 1774 ++--- deps/LZMA-SDK/C/Util/7z/Precomp.c | 8 +- deps/LZMA-SDK/C/Util/7z/Precomp.h | 20 +- deps/LZMA-SDK/C/Util/7z/makefile | 80 +- deps/LZMA-SDK/C/Util/7z/makefile.gcc | 68 +- deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.c | 572 +- deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsp | 344 +- deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsw | 58 +- deps/LZMA-SDK/C/Util/Lzma/makefile | 60 +- deps/LZMA-SDK/C/Util/Lzma/makefile.gcc | 42 +- deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.def | 8 +- deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsp | 364 +- deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsw | 58 +- deps/LZMA-SDK/C/Util/LzmaLib/LzmaLibExports.c | 28 +- deps/LZMA-SDK/C/Util/LzmaLib/makefile | 72 +- deps/LZMA-SDK/C/Util/LzmaLib/resource.rc | 6 +- deps/LZMA-SDK/C/Util/SfxSetup/Precomp.c | 8 +- deps/LZMA-SDK/C/Util/SfxSetup/Precomp.h | 20 +- deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.c | 1280 ++-- deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.dsp | 462 +- deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.dsw | 58 +- deps/LZMA-SDK/C/Util/SfxSetup/makefile | 74 +- deps/LZMA-SDK/C/Util/SfxSetup/makefile_con | 76 +- deps/LZMA-SDK/C/Util/SfxSetup/resource.rc | 10 +- deps/LZMA-SDK/C/Xz.c | 180 +- deps/LZMA-SDK/C/Xz.h | 1034 +-- deps/LZMA-SDK/C/XzCrc64.c | 172 +- deps/LZMA-SDK/C/XzCrc64.h | 52 +- deps/LZMA-SDK/C/XzCrc64Opt.c | 142 +- deps/LZMA-SDK/C/XzDec.c | 5674 +++++++-------- deps/LZMA-SDK/C/XzEnc.c | 2660 +++---- deps/LZMA-SDK/C/XzEnc.h | 120 +- deps/LZMA-SDK/C/XzIn.c | 650 +- deps/LZMA-SDK/C/var_clang.mak | 22 +- deps/LZMA-SDK/C/var_clang_arm64.mak | 22 +- deps/LZMA-SDK/C/var_clang_x64.mak | 22 +- deps/LZMA-SDK/C/var_clang_x86.mak | 22 +- deps/LZMA-SDK/C/var_gcc.mak | 24 +- deps/LZMA-SDK/C/var_gcc_arm64.mak | 24 +- deps/LZMA-SDK/C/var_gcc_x64.mak | 20 +- deps/LZMA-SDK/C/var_gcc_x86.mak | 20 +- deps/LZMA-SDK/C/var_mac_arm64.mak | 22 +- deps/LZMA-SDK/C/var_mac_x64.mak | 22 +- deps/LZMA-SDK/C/warn_clang.mak | 74 +- deps/LZMA-SDK/C/warn_clang_mac.mak | 74 +- deps/LZMA-SDK/C/warn_gcc.mak | 102 +- deps/LZMA-SDK/DOC/7zC.txt | 374 +- deps/LZMA-SDK/DOC/7zFormat.txt | 938 +-- deps/LZMA-SDK/DOC/Methods.txt | 346 +- deps/LZMA-SDK/DOC/installer.txt | 332 +- deps/LZMA-SDK/DOC/lzma-history.txt | 1040 +-- deps/LZMA-SDK/DOC/lzma-sdk.txt | 808 +-- deps/LZMA-SDK/DOC/lzma-specification.txt | 2352 +++--- deps/LZMA-SDK/DOC/lzma.txt | 656 +- 132 files changed, 39006 insertions(+), 39006 deletions(-) diff --git a/deps/LZMA-SDK/C/7z.h b/deps/LZMA-SDK/C/7z.h index 969523cd3..304f75ffc 100644 --- a/deps/LZMA-SDK/C/7z.h +++ b/deps/LZMA-SDK/C/7z.h @@ -1,204 +1,204 @@ -/* 7z.h -- 7z interface -2018-07-02 : Igor Pavlov : Public domain */ - -#ifndef __7Z_H -#define __7Z_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -#define k7zStartHeaderSize 0x20 -#define k7zSignatureSize 6 - -extern const Byte k7zSignature[k7zSignatureSize]; - -typedef struct -{ - const Byte *Data; - size_t Size; -} CSzData; - -/* CSzCoderInfo & CSzFolder support only default methods */ - -typedef struct -{ - size_t PropsOffset; - UInt32 MethodID; - Byte NumStreams; - Byte PropsSize; -} CSzCoderInfo; - -typedef struct -{ - UInt32 InIndex; - UInt32 OutIndex; -} CSzBond; - -#define SZ_NUM_CODERS_IN_FOLDER_MAX 4 -#define SZ_NUM_BONDS_IN_FOLDER_MAX 3 -#define SZ_NUM_PACK_STREAMS_IN_FOLDER_MAX 4 - -typedef struct -{ - UInt32 NumCoders; - UInt32 NumBonds; - UInt32 NumPackStreams; - UInt32 UnpackStream; - UInt32 PackStreams[SZ_NUM_PACK_STREAMS_IN_FOLDER_MAX]; - CSzBond Bonds[SZ_NUM_BONDS_IN_FOLDER_MAX]; - CSzCoderInfo Coders[SZ_NUM_CODERS_IN_FOLDER_MAX]; -} CSzFolder; - - -SRes SzGetNextFolderItem(CSzFolder *f, CSzData *sd); - -typedef struct -{ - UInt32 Low; - UInt32 High; -} CNtfsFileTime; - -typedef struct -{ - Byte *Defs; /* MSB 0 bit numbering */ - UInt32 *Vals; -} CSzBitUi32s; - -typedef struct -{ - Byte *Defs; /* MSB 0 bit numbering */ - // UInt64 *Vals; - CNtfsFileTime *Vals; -} CSzBitUi64s; - -#define SzBitArray_Check(p, i) (((p)[(i) >> 3] & (0x80 >> ((i) & 7))) != 0) - -#define SzBitWithVals_Check(p, i) ((p)->Defs && ((p)->Defs[(i) >> 3] & (0x80 >> ((i) & 7))) != 0) - -typedef struct -{ - UInt32 NumPackStreams; - UInt32 NumFolders; - - UInt64 *PackPositions; // NumPackStreams + 1 - CSzBitUi32s FolderCRCs; // NumFolders - - size_t *FoCodersOffsets; // NumFolders + 1 - UInt32 *FoStartPackStreamIndex; // NumFolders + 1 - UInt32 *FoToCoderUnpackSizes; // NumFolders + 1 - Byte *FoToMainUnpackSizeIndex; // NumFolders - UInt64 *CoderUnpackSizes; // for all coders in all folders - - Byte *CodersData; - - UInt64 RangeLimit; -} CSzAr; - -UInt64 SzAr_GetFolderUnpackSize(const CSzAr *p, UInt32 folderIndex); - -SRes SzAr_DecodeFolder(const CSzAr *p, UInt32 folderIndex, - ILookInStream *stream, UInt64 startPos, - Byte *outBuffer, size_t outSize, - ISzAllocPtr allocMain); - -typedef struct -{ - CSzAr db; - - UInt64 startPosAfterHeader; - UInt64 dataPos; - - UInt32 NumFiles; - - UInt64 *UnpackPositions; // NumFiles + 1 - // Byte *IsEmptyFiles; - Byte *IsDirs; - CSzBitUi32s CRCs; - - CSzBitUi32s Attribs; - // CSzBitUi32s Parents; - CSzBitUi64s MTime; - CSzBitUi64s CTime; - - UInt32 *FolderToFile; // NumFolders + 1 - UInt32 *FileToFolder; // NumFiles - - size_t *FileNameOffsets; /* in 2-byte steps */ - Byte *FileNames; /* UTF-16-LE */ -} CSzArEx; - -#define SzArEx_IsDir(p, i) (SzBitArray_Check((p)->IsDirs, i)) - -#define SzArEx_GetFileSize(p, i) ((p)->UnpackPositions[(i) + 1] - (p)->UnpackPositions[i]) - -void SzArEx_Init(CSzArEx *p); -void SzArEx_Free(CSzArEx *p, ISzAllocPtr alloc); -UInt64 SzArEx_GetFolderStreamPos(const CSzArEx *p, UInt32 folderIndex, UInt32 indexInFolder); -int SzArEx_GetFolderFullPackSize(const CSzArEx *p, UInt32 folderIndex, UInt64 *resSize); - -/* -if dest == NULL, the return value specifies the required size of the buffer, - in 16-bit characters, including the null-terminating character. -if dest != NULL, the return value specifies the number of 16-bit characters that - are written to the dest, including the null-terminating character. */ - -size_t SzArEx_GetFileNameUtf16(const CSzArEx *p, size_t fileIndex, UInt16 *dest); - -/* -size_t SzArEx_GetFullNameLen(const CSzArEx *p, size_t fileIndex); -UInt16 *SzArEx_GetFullNameUtf16_Back(const CSzArEx *p, size_t fileIndex, UInt16 *dest); -*/ - - - -/* - SzArEx_Extract extracts file from archive - - *outBuffer must be 0 before first call for each new archive. - - Extracting cache: - If you need to decompress more than one file, you can send - these values from previous call: - *blockIndex, - *outBuffer, - *outBufferSize - You can consider "*outBuffer" as cache of solid block. If your archive is solid, - it will increase decompression speed. - - If you use external function, you can declare these 3 cache variables - (blockIndex, outBuffer, outBufferSize) as static in that external function. - - Free *outBuffer and set *outBuffer to 0, if you want to flush cache. -*/ - -SRes SzArEx_Extract( - const CSzArEx *db, - ILookInStream *inStream, - UInt32 fileIndex, /* index of file */ - UInt32 *blockIndex, /* index of solid block */ - Byte **outBuffer, /* pointer to pointer to output buffer (allocated with allocMain) */ - size_t *outBufferSize, /* buffer size for output buffer */ - size_t *offset, /* offset of stream for required file in *outBuffer */ - size_t *outSizeProcessed, /* size of file in *outBuffer */ - ISzAllocPtr allocMain, - ISzAllocPtr allocTemp); - - -/* -SzArEx_Open Errors: -SZ_ERROR_NO_ARCHIVE -SZ_ERROR_ARCHIVE -SZ_ERROR_UNSUPPORTED -SZ_ERROR_MEM -SZ_ERROR_CRC -SZ_ERROR_INPUT_EOF -SZ_ERROR_FAIL -*/ - -SRes SzArEx_Open(CSzArEx *p, ILookInStream *inStream, - ISzAllocPtr allocMain, ISzAllocPtr allocTemp); - -EXTERN_C_END - -#endif +/* 7z.h -- 7z interface +2018-07-02 : Igor Pavlov : Public domain */ + +#ifndef __7Z_H +#define __7Z_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define k7zStartHeaderSize 0x20 +#define k7zSignatureSize 6 + +extern const Byte k7zSignature[k7zSignatureSize]; + +typedef struct +{ + const Byte *Data; + size_t Size; +} CSzData; + +/* CSzCoderInfo & CSzFolder support only default methods */ + +typedef struct +{ + size_t PropsOffset; + UInt32 MethodID; + Byte NumStreams; + Byte PropsSize; +} CSzCoderInfo; + +typedef struct +{ + UInt32 InIndex; + UInt32 OutIndex; +} CSzBond; + +#define SZ_NUM_CODERS_IN_FOLDER_MAX 4 +#define SZ_NUM_BONDS_IN_FOLDER_MAX 3 +#define SZ_NUM_PACK_STREAMS_IN_FOLDER_MAX 4 + +typedef struct +{ + UInt32 NumCoders; + UInt32 NumBonds; + UInt32 NumPackStreams; + UInt32 UnpackStream; + UInt32 PackStreams[SZ_NUM_PACK_STREAMS_IN_FOLDER_MAX]; + CSzBond Bonds[SZ_NUM_BONDS_IN_FOLDER_MAX]; + CSzCoderInfo Coders[SZ_NUM_CODERS_IN_FOLDER_MAX]; +} CSzFolder; + + +SRes SzGetNextFolderItem(CSzFolder *f, CSzData *sd); + +typedef struct +{ + UInt32 Low; + UInt32 High; +} CNtfsFileTime; + +typedef struct +{ + Byte *Defs; /* MSB 0 bit numbering */ + UInt32 *Vals; +} CSzBitUi32s; + +typedef struct +{ + Byte *Defs; /* MSB 0 bit numbering */ + // UInt64 *Vals; + CNtfsFileTime *Vals; +} CSzBitUi64s; + +#define SzBitArray_Check(p, i) (((p)[(i) >> 3] & (0x80 >> ((i) & 7))) != 0) + +#define SzBitWithVals_Check(p, i) ((p)->Defs && ((p)->Defs[(i) >> 3] & (0x80 >> ((i) & 7))) != 0) + +typedef struct +{ + UInt32 NumPackStreams; + UInt32 NumFolders; + + UInt64 *PackPositions; // NumPackStreams + 1 + CSzBitUi32s FolderCRCs; // NumFolders + + size_t *FoCodersOffsets; // NumFolders + 1 + UInt32 *FoStartPackStreamIndex; // NumFolders + 1 + UInt32 *FoToCoderUnpackSizes; // NumFolders + 1 + Byte *FoToMainUnpackSizeIndex; // NumFolders + UInt64 *CoderUnpackSizes; // for all coders in all folders + + Byte *CodersData; + + UInt64 RangeLimit; +} CSzAr; + +UInt64 SzAr_GetFolderUnpackSize(const CSzAr *p, UInt32 folderIndex); + +SRes SzAr_DecodeFolder(const CSzAr *p, UInt32 folderIndex, + ILookInStream *stream, UInt64 startPos, + Byte *outBuffer, size_t outSize, + ISzAllocPtr allocMain); + +typedef struct +{ + CSzAr db; + + UInt64 startPosAfterHeader; + UInt64 dataPos; + + UInt32 NumFiles; + + UInt64 *UnpackPositions; // NumFiles + 1 + // Byte *IsEmptyFiles; + Byte *IsDirs; + CSzBitUi32s CRCs; + + CSzBitUi32s Attribs; + // CSzBitUi32s Parents; + CSzBitUi64s MTime; + CSzBitUi64s CTime; + + UInt32 *FolderToFile; // NumFolders + 1 + UInt32 *FileToFolder; // NumFiles + + size_t *FileNameOffsets; /* in 2-byte steps */ + Byte *FileNames; /* UTF-16-LE */ +} CSzArEx; + +#define SzArEx_IsDir(p, i) (SzBitArray_Check((p)->IsDirs, i)) + +#define SzArEx_GetFileSize(p, i) ((p)->UnpackPositions[(i) + 1] - (p)->UnpackPositions[i]) + +void SzArEx_Init(CSzArEx *p); +void SzArEx_Free(CSzArEx *p, ISzAllocPtr alloc); +UInt64 SzArEx_GetFolderStreamPos(const CSzArEx *p, UInt32 folderIndex, UInt32 indexInFolder); +int SzArEx_GetFolderFullPackSize(const CSzArEx *p, UInt32 folderIndex, UInt64 *resSize); + +/* +if dest == NULL, the return value specifies the required size of the buffer, + in 16-bit characters, including the null-terminating character. +if dest != NULL, the return value specifies the number of 16-bit characters that + are written to the dest, including the null-terminating character. */ + +size_t SzArEx_GetFileNameUtf16(const CSzArEx *p, size_t fileIndex, UInt16 *dest); + +/* +size_t SzArEx_GetFullNameLen(const CSzArEx *p, size_t fileIndex); +UInt16 *SzArEx_GetFullNameUtf16_Back(const CSzArEx *p, size_t fileIndex, UInt16 *dest); +*/ + + + +/* + SzArEx_Extract extracts file from archive + + *outBuffer must be 0 before first call for each new archive. + + Extracting cache: + If you need to decompress more than one file, you can send + these values from previous call: + *blockIndex, + *outBuffer, + *outBufferSize + You can consider "*outBuffer" as cache of solid block. If your archive is solid, + it will increase decompression speed. + + If you use external function, you can declare these 3 cache variables + (blockIndex, outBuffer, outBufferSize) as static in that external function. + + Free *outBuffer and set *outBuffer to 0, if you want to flush cache. +*/ + +SRes SzArEx_Extract( + const CSzArEx *db, + ILookInStream *inStream, + UInt32 fileIndex, /* index of file */ + UInt32 *blockIndex, /* index of solid block */ + Byte **outBuffer, /* pointer to pointer to output buffer (allocated with allocMain) */ + size_t *outBufferSize, /* buffer size for output buffer */ + size_t *offset, /* offset of stream for required file in *outBuffer */ + size_t *outSizeProcessed, /* size of file in *outBuffer */ + ISzAllocPtr allocMain, + ISzAllocPtr allocTemp); + + +/* +SzArEx_Open Errors: +SZ_ERROR_NO_ARCHIVE +SZ_ERROR_ARCHIVE +SZ_ERROR_UNSUPPORTED +SZ_ERROR_MEM +SZ_ERROR_CRC +SZ_ERROR_INPUT_EOF +SZ_ERROR_FAIL +*/ + +SRes SzArEx_Open(CSzArEx *p, ILookInStream *inStream, + ISzAllocPtr allocMain, ISzAllocPtr allocTemp); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/7zAlloc.c b/deps/LZMA-SDK/C/7zAlloc.c index ea32809c6..c924a529f 100644 --- a/deps/LZMA-SDK/C/7zAlloc.c +++ b/deps/LZMA-SDK/C/7zAlloc.c @@ -1,80 +1,80 @@ -/* 7zAlloc.c -- Allocation functions -2017-04-03 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include - -#include "7zAlloc.h" - -/* #define _SZ_ALLOC_DEBUG */ -/* use _SZ_ALLOC_DEBUG to debug alloc/free operations */ - -#ifdef _SZ_ALLOC_DEBUG - -#ifdef _WIN32 -#include -#endif - -#include -int g_allocCount = 0; -int g_allocCountTemp = 0; - -#endif - -void *SzAlloc(ISzAllocPtr p, size_t size) -{ - UNUSED_VAR(p); - if (size == 0) - return 0; - #ifdef _SZ_ALLOC_DEBUG - fprintf(stderr, "\nAlloc %10u bytes; count = %10d", (unsigned)size, g_allocCount); - g_allocCount++; - #endif - return malloc(size); -} - -void SzFree(ISzAllocPtr p, void *address) -{ - UNUSED_VAR(p); - #ifdef _SZ_ALLOC_DEBUG - if (address != 0) - { - g_allocCount--; - fprintf(stderr, "\nFree; count = %10d", g_allocCount); - } - #endif - free(address); -} - -void *SzAllocTemp(ISzAllocPtr p, size_t size) -{ - UNUSED_VAR(p); - if (size == 0) - return 0; - #ifdef _SZ_ALLOC_DEBUG - fprintf(stderr, "\nAlloc_temp %10u bytes; count = %10d", (unsigned)size, g_allocCountTemp); - g_allocCountTemp++; - #ifdef _WIN32 - return HeapAlloc(GetProcessHeap(), 0, size); - #endif - #endif - return malloc(size); -} - -void SzFreeTemp(ISzAllocPtr p, void *address) -{ - UNUSED_VAR(p); - #ifdef _SZ_ALLOC_DEBUG - if (address != 0) - { - g_allocCountTemp--; - fprintf(stderr, "\nFree_temp; count = %10d", g_allocCountTemp); - } - #ifdef _WIN32 - HeapFree(GetProcessHeap(), 0, address); - return; - #endif - #endif - free(address); -} +/* 7zAlloc.c -- Allocation functions +2017-04-03 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#include "7zAlloc.h" + +/* #define _SZ_ALLOC_DEBUG */ +/* use _SZ_ALLOC_DEBUG to debug alloc/free operations */ + +#ifdef _SZ_ALLOC_DEBUG + +#ifdef _WIN32 +#include +#endif + +#include +int g_allocCount = 0; +int g_allocCountTemp = 0; + +#endif + +void *SzAlloc(ISzAllocPtr p, size_t size) +{ + UNUSED_VAR(p); + if (size == 0) + return 0; + #ifdef _SZ_ALLOC_DEBUG + fprintf(stderr, "\nAlloc %10u bytes; count = %10d", (unsigned)size, g_allocCount); + g_allocCount++; + #endif + return malloc(size); +} + +void SzFree(ISzAllocPtr p, void *address) +{ + UNUSED_VAR(p); + #ifdef _SZ_ALLOC_DEBUG + if (address != 0) + { + g_allocCount--; + fprintf(stderr, "\nFree; count = %10d", g_allocCount); + } + #endif + free(address); +} + +void *SzAllocTemp(ISzAllocPtr p, size_t size) +{ + UNUSED_VAR(p); + if (size == 0) + return 0; + #ifdef _SZ_ALLOC_DEBUG + fprintf(stderr, "\nAlloc_temp %10u bytes; count = %10d", (unsigned)size, g_allocCountTemp); + g_allocCountTemp++; + #ifdef _WIN32 + return HeapAlloc(GetProcessHeap(), 0, size); + #endif + #endif + return malloc(size); +} + +void SzFreeTemp(ISzAllocPtr p, void *address) +{ + UNUSED_VAR(p); + #ifdef _SZ_ALLOC_DEBUG + if (address != 0) + { + g_allocCountTemp--; + fprintf(stderr, "\nFree_temp; count = %10d", g_allocCountTemp); + } + #ifdef _WIN32 + HeapFree(GetProcessHeap(), 0, address); + return; + #endif + #endif + free(address); +} diff --git a/deps/LZMA-SDK/C/7zAlloc.h b/deps/LZMA-SDK/C/7zAlloc.h index c0f89d73c..44778f9b2 100644 --- a/deps/LZMA-SDK/C/7zAlloc.h +++ b/deps/LZMA-SDK/C/7zAlloc.h @@ -1,19 +1,19 @@ -/* 7zAlloc.h -- Allocation functions -2017-04-03 : Igor Pavlov : Public domain */ - -#ifndef __7Z_ALLOC_H -#define __7Z_ALLOC_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -void *SzAlloc(ISzAllocPtr p, size_t size); -void SzFree(ISzAllocPtr p, void *address); - -void *SzAllocTemp(ISzAllocPtr p, size_t size); -void SzFreeTemp(ISzAllocPtr p, void *address); - -EXTERN_C_END - -#endif +/* 7zAlloc.h -- Allocation functions +2017-04-03 : Igor Pavlov : Public domain */ + +#ifndef __7Z_ALLOC_H +#define __7Z_ALLOC_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +void *SzAlloc(ISzAllocPtr p, size_t size); +void SzFree(ISzAllocPtr p, void *address); + +void *SzAllocTemp(ISzAllocPtr p, size_t size); +void SzFreeTemp(ISzAllocPtr p, void *address); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/7zArcIn.c b/deps/LZMA-SDK/C/7zArcIn.c index 7ccc72101..0d9dec41e 100644 --- a/deps/LZMA-SDK/C/7zArcIn.c +++ b/deps/LZMA-SDK/C/7zArcIn.c @@ -1,1783 +1,1783 @@ -/* 7zArcIn.c -- 7z Input functions -2021-02-09 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include - -#include "7z.h" -#include "7zBuf.h" -#include "7zCrc.h" -#include "CpuArch.h" - -#define MY_ALLOC(T, p, size, alloc) { \ - if ((p = (T *)ISzAlloc_Alloc(alloc, (size) * sizeof(T))) == NULL) return SZ_ERROR_MEM; } - -#define MY_ALLOC_ZE(T, p, size, alloc) { if ((size) == 0) p = NULL; else MY_ALLOC(T, p, size, alloc) } - -#define MY_ALLOC_AND_CPY(to, size, from, alloc) \ - { MY_ALLOC(Byte, to, size, alloc); memcpy(to, from, size); } - -#define MY_ALLOC_ZE_AND_CPY(to, size, from, alloc) \ - { if ((size) == 0) to = NULL; else { MY_ALLOC_AND_CPY(to, size, from, alloc) } } - -#define k7zMajorVersion 0 - -enum EIdEnum -{ - k7zIdEnd, - k7zIdHeader, - k7zIdArchiveProperties, - k7zIdAdditionalStreamsInfo, - k7zIdMainStreamsInfo, - k7zIdFilesInfo, - k7zIdPackInfo, - k7zIdUnpackInfo, - k7zIdSubStreamsInfo, - k7zIdSize, - k7zIdCRC, - k7zIdFolder, - k7zIdCodersUnpackSize, - k7zIdNumUnpackStream, - k7zIdEmptyStream, - k7zIdEmptyFile, - k7zIdAnti, - k7zIdName, - k7zIdCTime, - k7zIdATime, - k7zIdMTime, - k7zIdWinAttrib, - k7zIdComment, - k7zIdEncodedHeader, - k7zIdStartPos, - k7zIdDummy - // k7zNtSecure, - // k7zParent, - // k7zIsReal -}; - -const Byte k7zSignature[k7zSignatureSize] = {'7', 'z', 0xBC, 0xAF, 0x27, 0x1C}; - -#define SzBitUi32s_Init(p) { (p)->Defs = NULL; (p)->Vals = NULL; } - -static SRes SzBitUi32s_Alloc(CSzBitUi32s *p, size_t num, ISzAllocPtr alloc) -{ - if (num == 0) - { - p->Defs = NULL; - p->Vals = NULL; - } - else - { - MY_ALLOC(Byte, p->Defs, (num + 7) >> 3, alloc); - MY_ALLOC(UInt32, p->Vals, num, alloc); - } - return SZ_OK; -} - -static void SzBitUi32s_Free(CSzBitUi32s *p, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, p->Defs); p->Defs = NULL; - ISzAlloc_Free(alloc, p->Vals); p->Vals = NULL; -} - -#define SzBitUi64s_Init(p) { (p)->Defs = NULL; (p)->Vals = NULL; } - -static void SzBitUi64s_Free(CSzBitUi64s *p, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, p->Defs); p->Defs = NULL; - ISzAlloc_Free(alloc, p->Vals); p->Vals = NULL; -} - - -static void SzAr_Init(CSzAr *p) -{ - p->NumPackStreams = 0; - p->NumFolders = 0; - - p->PackPositions = NULL; - SzBitUi32s_Init(&p->FolderCRCs); - - p->FoCodersOffsets = NULL; - p->FoStartPackStreamIndex = NULL; - p->FoToCoderUnpackSizes = NULL; - p->FoToMainUnpackSizeIndex = NULL; - p->CoderUnpackSizes = NULL; - - p->CodersData = NULL; - - p->RangeLimit = 0; -} - -static void SzAr_Free(CSzAr *p, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, p->PackPositions); - SzBitUi32s_Free(&p->FolderCRCs, alloc); - - ISzAlloc_Free(alloc, p->FoCodersOffsets); - ISzAlloc_Free(alloc, p->FoStartPackStreamIndex); - ISzAlloc_Free(alloc, p->FoToCoderUnpackSizes); - ISzAlloc_Free(alloc, p->FoToMainUnpackSizeIndex); - ISzAlloc_Free(alloc, p->CoderUnpackSizes); - - ISzAlloc_Free(alloc, p->CodersData); - - SzAr_Init(p); -} - - -void SzArEx_Init(CSzArEx *p) -{ - SzAr_Init(&p->db); - - p->NumFiles = 0; - p->dataPos = 0; - - p->UnpackPositions = NULL; - p->IsDirs = NULL; - - p->FolderToFile = NULL; - p->FileToFolder = NULL; - - p->FileNameOffsets = NULL; - p->FileNames = NULL; - - SzBitUi32s_Init(&p->CRCs); - SzBitUi32s_Init(&p->Attribs); - // SzBitUi32s_Init(&p->Parents); - SzBitUi64s_Init(&p->MTime); - SzBitUi64s_Init(&p->CTime); -} - -void SzArEx_Free(CSzArEx *p, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, p->UnpackPositions); - ISzAlloc_Free(alloc, p->IsDirs); - - ISzAlloc_Free(alloc, p->FolderToFile); - ISzAlloc_Free(alloc, p->FileToFolder); - - ISzAlloc_Free(alloc, p->FileNameOffsets); - ISzAlloc_Free(alloc, p->FileNames); - - SzBitUi32s_Free(&p->CRCs, alloc); - SzBitUi32s_Free(&p->Attribs, alloc); - // SzBitUi32s_Free(&p->Parents, alloc); - SzBitUi64s_Free(&p->MTime, alloc); - SzBitUi64s_Free(&p->CTime, alloc); - - SzAr_Free(&p->db, alloc); - SzArEx_Init(p); -} - - -static int TestSignatureCandidate(const Byte *testBytes) -{ - unsigned i; - for (i = 0; i < k7zSignatureSize; i++) - if (testBytes[i] != k7zSignature[i]) - return 0; - return 1; -} - -#define SzData_Clear(p) { (p)->Data = NULL; (p)->Size = 0; } - -#define SZ_READ_BYTE_SD(_sd_, dest) if ((_sd_)->Size == 0) return SZ_ERROR_ARCHIVE; (_sd_)->Size--; dest = *(_sd_)->Data++; -#define SZ_READ_BYTE(dest) SZ_READ_BYTE_SD(sd, dest) -#define SZ_READ_BYTE_2(dest) if (sd.Size == 0) return SZ_ERROR_ARCHIVE; sd.Size--; dest = *sd.Data++; - -#define SKIP_DATA(sd, size) { sd->Size -= (size_t)(size); sd->Data += (size_t)(size); } -#define SKIP_DATA2(sd, size) { sd.Size -= (size_t)(size); sd.Data += (size_t)(size); } - -#define SZ_READ_32(dest) if (sd.Size < 4) return SZ_ERROR_ARCHIVE; \ - dest = GetUi32(sd.Data); SKIP_DATA2(sd, 4); - -static MY_NO_INLINE SRes ReadNumber(CSzData *sd, UInt64 *value) -{ - Byte firstByte, mask; - unsigned i; - UInt32 v; - - SZ_READ_BYTE(firstByte); - if ((firstByte & 0x80) == 0) - { - *value = firstByte; - return SZ_OK; - } - SZ_READ_BYTE(v); - if ((firstByte & 0x40) == 0) - { - *value = (((UInt32)firstByte & 0x3F) << 8) | v; - return SZ_OK; - } - SZ_READ_BYTE(mask); - *value = v | ((UInt32)mask << 8); - mask = 0x20; - for (i = 2; i < 8; i++) - { - Byte b; - if ((firstByte & mask) == 0) - { - UInt64 highPart = (unsigned)firstByte & (unsigned)(mask - 1); - *value |= (highPart << (8 * i)); - return SZ_OK; - } - SZ_READ_BYTE(b); - *value |= ((UInt64)b << (8 * i)); - mask >>= 1; - } - return SZ_OK; -} - - -static MY_NO_INLINE SRes SzReadNumber32(CSzData *sd, UInt32 *value) -{ - Byte firstByte; - UInt64 value64; - if (sd->Size == 0) - return SZ_ERROR_ARCHIVE; - firstByte = *sd->Data; - if ((firstByte & 0x80) == 0) - { - *value = firstByte; - sd->Data++; - sd->Size--; - return SZ_OK; - } - RINOK(ReadNumber(sd, &value64)); - if (value64 >= (UInt32)0x80000000 - 1) - return SZ_ERROR_UNSUPPORTED; - if (value64 >= ((UInt64)(1) << ((sizeof(size_t) - 1) * 8 + 4))) - return SZ_ERROR_UNSUPPORTED; - *value = (UInt32)value64; - return SZ_OK; -} - -#define ReadID(sd, value) ReadNumber(sd, value) - -static SRes SkipData(CSzData *sd) -{ - UInt64 size; - RINOK(ReadNumber(sd, &size)); - if (size > sd->Size) - return SZ_ERROR_ARCHIVE; - SKIP_DATA(sd, size); - return SZ_OK; -} - -static SRes WaitId(CSzData *sd, UInt32 id) -{ - for (;;) - { - UInt64 type; - RINOK(ReadID(sd, &type)); - if (type == id) - return SZ_OK; - if (type == k7zIdEnd) - return SZ_ERROR_ARCHIVE; - RINOK(SkipData(sd)); - } -} - -static SRes RememberBitVector(CSzData *sd, UInt32 numItems, const Byte **v) -{ - UInt32 numBytes = (numItems + 7) >> 3; - if (numBytes > sd->Size) - return SZ_ERROR_ARCHIVE; - *v = sd->Data; - SKIP_DATA(sd, numBytes); - return SZ_OK; -} - -static UInt32 CountDefinedBits(const Byte *bits, UInt32 numItems) -{ - Byte b = 0; - unsigned m = 0; - UInt32 sum = 0; - for (; numItems != 0; numItems--) - { - if (m == 0) - { - b = *bits++; - m = 8; - } - m--; - sum += ((b >> m) & 1); - } - return sum; -} - -static MY_NO_INLINE SRes ReadBitVector(CSzData *sd, UInt32 numItems, Byte **v, ISzAllocPtr alloc) -{ - Byte allAreDefined; - Byte *v2; - UInt32 numBytes = (numItems + 7) >> 3; - *v = NULL; - SZ_READ_BYTE(allAreDefined); - if (numBytes == 0) - return SZ_OK; - if (allAreDefined == 0) - { - if (numBytes > sd->Size) - return SZ_ERROR_ARCHIVE; - MY_ALLOC_AND_CPY(*v, numBytes, sd->Data, alloc); - SKIP_DATA(sd, numBytes); - return SZ_OK; - } - MY_ALLOC(Byte, *v, numBytes, alloc); - v2 = *v; - memset(v2, 0xFF, (size_t)numBytes); - { - unsigned numBits = (unsigned)numItems & 7; - if (numBits != 0) - v2[(size_t)numBytes - 1] = (Byte)((((UInt32)1 << numBits) - 1) << (8 - numBits)); - } - return SZ_OK; -} - -static MY_NO_INLINE SRes ReadUi32s(CSzData *sd2, UInt32 numItems, CSzBitUi32s *crcs, ISzAllocPtr alloc) -{ - UInt32 i; - CSzData sd; - UInt32 *vals; - const Byte *defs; - MY_ALLOC_ZE(UInt32, crcs->Vals, numItems, alloc); - sd = *sd2; - defs = crcs->Defs; - vals = crcs->Vals; - for (i = 0; i < numItems; i++) - if (SzBitArray_Check(defs, i)) - { - SZ_READ_32(vals[i]); - } - else - vals[i] = 0; - *sd2 = sd; - return SZ_OK; -} - -static SRes ReadBitUi32s(CSzData *sd, UInt32 numItems, CSzBitUi32s *crcs, ISzAllocPtr alloc) -{ - SzBitUi32s_Free(crcs, alloc); - RINOK(ReadBitVector(sd, numItems, &crcs->Defs, alloc)); - return ReadUi32s(sd, numItems, crcs, alloc); -} - -static SRes SkipBitUi32s(CSzData *sd, UInt32 numItems) -{ - Byte allAreDefined; - UInt32 numDefined = numItems; - SZ_READ_BYTE(allAreDefined); - if (!allAreDefined) - { - size_t numBytes = (numItems + 7) >> 3; - if (numBytes > sd->Size) - return SZ_ERROR_ARCHIVE; - numDefined = CountDefinedBits(sd->Data, numItems); - SKIP_DATA(sd, numBytes); - } - if (numDefined > (sd->Size >> 2)) - return SZ_ERROR_ARCHIVE; - SKIP_DATA(sd, (size_t)numDefined * 4); - return SZ_OK; -} - -static SRes ReadPackInfo(CSzAr *p, CSzData *sd, ISzAllocPtr alloc) -{ - RINOK(SzReadNumber32(sd, &p->NumPackStreams)); - - RINOK(WaitId(sd, k7zIdSize)); - MY_ALLOC(UInt64, p->PackPositions, (size_t)p->NumPackStreams + 1, alloc); - { - UInt64 sum = 0; - UInt32 i; - UInt32 numPackStreams = p->NumPackStreams; - for (i = 0; i < numPackStreams; i++) - { - UInt64 packSize; - p->PackPositions[i] = sum; - RINOK(ReadNumber(sd, &packSize)); - sum += packSize; - if (sum < packSize) - return SZ_ERROR_ARCHIVE; - } - p->PackPositions[i] = sum; - } - - for (;;) - { - UInt64 type; - RINOK(ReadID(sd, &type)); - if (type == k7zIdEnd) - return SZ_OK; - if (type == k7zIdCRC) - { - /* CRC of packed streams is unused now */ - RINOK(SkipBitUi32s(sd, p->NumPackStreams)); - continue; - } - RINOK(SkipData(sd)); - } -} - -/* -static SRes SzReadSwitch(CSzData *sd) -{ - Byte external; - RINOK(SzReadByte(sd, &external)); - return (external == 0) ? SZ_OK: SZ_ERROR_UNSUPPORTED; -} -*/ - -#define k_NumCodersStreams_in_Folder_MAX (SZ_NUM_BONDS_IN_FOLDER_MAX + SZ_NUM_PACK_STREAMS_IN_FOLDER_MAX) - -SRes SzGetNextFolderItem(CSzFolder *f, CSzData *sd) -{ - UInt32 numCoders, i; - UInt32 numInStreams = 0; - const Byte *dataStart = sd->Data; - - f->NumCoders = 0; - f->NumBonds = 0; - f->NumPackStreams = 0; - f->UnpackStream = 0; - - RINOK(SzReadNumber32(sd, &numCoders)); - if (numCoders == 0 || numCoders > SZ_NUM_CODERS_IN_FOLDER_MAX) - return SZ_ERROR_UNSUPPORTED; - - for (i = 0; i < numCoders; i++) - { - Byte mainByte; - CSzCoderInfo *coder = f->Coders + i; - unsigned idSize, j; - UInt64 id; - - SZ_READ_BYTE(mainByte); - if ((mainByte & 0xC0) != 0) - return SZ_ERROR_UNSUPPORTED; - - idSize = (unsigned)(mainByte & 0xF); - if (idSize > sizeof(id)) - return SZ_ERROR_UNSUPPORTED; - if (idSize > sd->Size) - return SZ_ERROR_ARCHIVE; - id = 0; - for (j = 0; j < idSize; j++) - { - id = ((id << 8) | *sd->Data); - sd->Data++; - sd->Size--; - } - if (id > (UInt32)0xFFFFFFFF) - return SZ_ERROR_UNSUPPORTED; - coder->MethodID = (UInt32)id; - - coder->NumStreams = 1; - coder->PropsOffset = 0; - coder->PropsSize = 0; - - if ((mainByte & 0x10) != 0) - { - UInt32 numStreams; - - RINOK(SzReadNumber32(sd, &numStreams)); - if (numStreams > k_NumCodersStreams_in_Folder_MAX) - return SZ_ERROR_UNSUPPORTED; - coder->NumStreams = (Byte)numStreams; - - RINOK(SzReadNumber32(sd, &numStreams)); - if (numStreams != 1) - return SZ_ERROR_UNSUPPORTED; - } - - numInStreams += coder->NumStreams; - - if (numInStreams > k_NumCodersStreams_in_Folder_MAX) - return SZ_ERROR_UNSUPPORTED; - - if ((mainByte & 0x20) != 0) - { - UInt32 propsSize = 0; - RINOK(SzReadNumber32(sd, &propsSize)); - if (propsSize > sd->Size) - return SZ_ERROR_ARCHIVE; - if (propsSize >= 0x80) - return SZ_ERROR_UNSUPPORTED; - coder->PropsOffset = (size_t)(sd->Data - dataStart); - coder->PropsSize = (Byte)propsSize; - sd->Data += (size_t)propsSize; - sd->Size -= (size_t)propsSize; - } - } - - /* - if (numInStreams == 1 && numCoders == 1) - { - f->NumPackStreams = 1; - f->PackStreams[0] = 0; - } - else - */ - { - Byte streamUsed[k_NumCodersStreams_in_Folder_MAX]; - UInt32 numBonds, numPackStreams; - - numBonds = numCoders - 1; - if (numInStreams < numBonds) - return SZ_ERROR_ARCHIVE; - if (numBonds > SZ_NUM_BONDS_IN_FOLDER_MAX) - return SZ_ERROR_UNSUPPORTED; - f->NumBonds = numBonds; - - numPackStreams = numInStreams - numBonds; - if (numPackStreams > SZ_NUM_PACK_STREAMS_IN_FOLDER_MAX) - return SZ_ERROR_UNSUPPORTED; - f->NumPackStreams = numPackStreams; - - for (i = 0; i < numInStreams; i++) - streamUsed[i] = False; - - if (numBonds != 0) - { - Byte coderUsed[SZ_NUM_CODERS_IN_FOLDER_MAX]; - - for (i = 0; i < numCoders; i++) - coderUsed[i] = False; - - for (i = 0; i < numBonds; i++) - { - CSzBond *bp = f->Bonds + i; - - RINOK(SzReadNumber32(sd, &bp->InIndex)); - if (bp->InIndex >= numInStreams || streamUsed[bp->InIndex]) - return SZ_ERROR_ARCHIVE; - streamUsed[bp->InIndex] = True; - - RINOK(SzReadNumber32(sd, &bp->OutIndex)); - if (bp->OutIndex >= numCoders || coderUsed[bp->OutIndex]) - return SZ_ERROR_ARCHIVE; - coderUsed[bp->OutIndex] = True; - } - - for (i = 0; i < numCoders; i++) - if (!coderUsed[i]) - { - f->UnpackStream = i; - break; - } - - if (i == numCoders) - return SZ_ERROR_ARCHIVE; - } - - if (numPackStreams == 1) - { - for (i = 0; i < numInStreams; i++) - if (!streamUsed[i]) - break; - if (i == numInStreams) - return SZ_ERROR_ARCHIVE; - f->PackStreams[0] = i; - } - else - for (i = 0; i < numPackStreams; i++) - { - UInt32 index; - RINOK(SzReadNumber32(sd, &index)); - if (index >= numInStreams || streamUsed[index]) - return SZ_ERROR_ARCHIVE; - streamUsed[index] = True; - f->PackStreams[i] = index; - } - } - - f->NumCoders = numCoders; - - return SZ_OK; -} - - -static MY_NO_INLINE SRes SkipNumbers(CSzData *sd2, UInt32 num) -{ - CSzData sd; - sd = *sd2; - for (; num != 0; num--) - { - Byte firstByte, mask; - unsigned i; - SZ_READ_BYTE_2(firstByte); - if ((firstByte & 0x80) == 0) - continue; - if ((firstByte & 0x40) == 0) - { - if (sd.Size == 0) - return SZ_ERROR_ARCHIVE; - sd.Size--; - sd.Data++; - continue; - } - mask = 0x20; - for (i = 2; i < 8 && (firstByte & mask) != 0; i++) - mask >>= 1; - if (i > sd.Size) - return SZ_ERROR_ARCHIVE; - SKIP_DATA2(sd, i); - } - *sd2 = sd; - return SZ_OK; -} - - -#define k_Scan_NumCoders_MAX 64 -#define k_Scan_NumCodersStreams_in_Folder_MAX 64 - - -static SRes ReadUnpackInfo(CSzAr *p, - CSzData *sd2, - UInt32 numFoldersMax, - const CBuf *tempBufs, UInt32 numTempBufs, - ISzAllocPtr alloc) -{ - CSzData sd; - - UInt32 fo, numFolders, numCodersOutStreams, packStreamIndex; - const Byte *startBufPtr; - Byte external; - - RINOK(WaitId(sd2, k7zIdFolder)); - - RINOK(SzReadNumber32(sd2, &numFolders)); - if (numFolders > numFoldersMax) - return SZ_ERROR_UNSUPPORTED; - p->NumFolders = numFolders; - - SZ_READ_BYTE_SD(sd2, external); - if (external == 0) - sd = *sd2; - else - { - UInt32 index; - RINOK(SzReadNumber32(sd2, &index)); - if (index >= numTempBufs) - return SZ_ERROR_ARCHIVE; - sd.Data = tempBufs[index].data; - sd.Size = tempBufs[index].size; - } - - MY_ALLOC(size_t, p->FoCodersOffsets, (size_t)numFolders + 1, alloc); - MY_ALLOC(UInt32, p->FoStartPackStreamIndex, (size_t)numFolders + 1, alloc); - MY_ALLOC(UInt32, p->FoToCoderUnpackSizes, (size_t)numFolders + 1, alloc); - MY_ALLOC_ZE(Byte, p->FoToMainUnpackSizeIndex, (size_t)numFolders, alloc); - - startBufPtr = sd.Data; - - packStreamIndex = 0; - numCodersOutStreams = 0; - - for (fo = 0; fo < numFolders; fo++) - { - UInt32 numCoders, ci, numInStreams = 0; - - p->FoCodersOffsets[fo] = (size_t)(sd.Data - startBufPtr); - - RINOK(SzReadNumber32(&sd, &numCoders)); - if (numCoders == 0 || numCoders > k_Scan_NumCoders_MAX) - return SZ_ERROR_UNSUPPORTED; - - for (ci = 0; ci < numCoders; ci++) - { - Byte mainByte; - unsigned idSize; - UInt32 coderInStreams; - - SZ_READ_BYTE_2(mainByte); - if ((mainByte & 0xC0) != 0) - return SZ_ERROR_UNSUPPORTED; - idSize = (mainByte & 0xF); - if (idSize > 8) - return SZ_ERROR_UNSUPPORTED; - if (idSize > sd.Size) - return SZ_ERROR_ARCHIVE; - SKIP_DATA2(sd, idSize); - - coderInStreams = 1; - - if ((mainByte & 0x10) != 0) - { - UInt32 coderOutStreams; - RINOK(SzReadNumber32(&sd, &coderInStreams)); - RINOK(SzReadNumber32(&sd, &coderOutStreams)); - if (coderInStreams > k_Scan_NumCodersStreams_in_Folder_MAX || coderOutStreams != 1) - return SZ_ERROR_UNSUPPORTED; - } - - numInStreams += coderInStreams; - - if ((mainByte & 0x20) != 0) - { - UInt32 propsSize; - RINOK(SzReadNumber32(&sd, &propsSize)); - if (propsSize > sd.Size) - return SZ_ERROR_ARCHIVE; - SKIP_DATA2(sd, propsSize); - } - } - - { - UInt32 indexOfMainStream = 0; - UInt32 numPackStreams = 1; - - if (numCoders != 1 || numInStreams != 1) - { - Byte streamUsed[k_Scan_NumCodersStreams_in_Folder_MAX]; - Byte coderUsed[k_Scan_NumCoders_MAX]; - - UInt32 i; - UInt32 numBonds = numCoders - 1; - if (numInStreams < numBonds) - return SZ_ERROR_ARCHIVE; - - if (numInStreams > k_Scan_NumCodersStreams_in_Folder_MAX) - return SZ_ERROR_UNSUPPORTED; - - for (i = 0; i < numInStreams; i++) - streamUsed[i] = False; - for (i = 0; i < numCoders; i++) - coderUsed[i] = False; - - for (i = 0; i < numBonds; i++) - { - UInt32 index; - - RINOK(SzReadNumber32(&sd, &index)); - if (index >= numInStreams || streamUsed[index]) - return SZ_ERROR_ARCHIVE; - streamUsed[index] = True; - - RINOK(SzReadNumber32(&sd, &index)); - if (index >= numCoders || coderUsed[index]) - return SZ_ERROR_ARCHIVE; - coderUsed[index] = True; - } - - numPackStreams = numInStreams - numBonds; - - if (numPackStreams != 1) - for (i = 0; i < numPackStreams; i++) - { - UInt32 index; - RINOK(SzReadNumber32(&sd, &index)); - if (index >= numInStreams || streamUsed[index]) - return SZ_ERROR_ARCHIVE; - streamUsed[index] = True; - } - - for (i = 0; i < numCoders; i++) - if (!coderUsed[i]) - { - indexOfMainStream = i; - break; - } - - if (i == numCoders) - return SZ_ERROR_ARCHIVE; - } - - p->FoStartPackStreamIndex[fo] = packStreamIndex; - p->FoToCoderUnpackSizes[fo] = numCodersOutStreams; - p->FoToMainUnpackSizeIndex[fo] = (Byte)indexOfMainStream; - numCodersOutStreams += numCoders; - if (numCodersOutStreams < numCoders) - return SZ_ERROR_UNSUPPORTED; - if (numPackStreams > p->NumPackStreams - packStreamIndex) - return SZ_ERROR_ARCHIVE; - packStreamIndex += numPackStreams; - } - } - - p->FoToCoderUnpackSizes[fo] = numCodersOutStreams; - - { - const size_t dataSize = (size_t)(sd.Data - startBufPtr); - p->FoStartPackStreamIndex[fo] = packStreamIndex; - p->FoCodersOffsets[fo] = dataSize; - MY_ALLOC_ZE_AND_CPY(p->CodersData, dataSize, startBufPtr, alloc); - } - - if (external != 0) - { - if (sd.Size != 0) - return SZ_ERROR_ARCHIVE; - sd = *sd2; - } - - RINOK(WaitId(&sd, k7zIdCodersUnpackSize)); - - MY_ALLOC_ZE(UInt64, p->CoderUnpackSizes, (size_t)numCodersOutStreams, alloc); - { - UInt32 i; - for (i = 0; i < numCodersOutStreams; i++) - { - RINOK(ReadNumber(&sd, p->CoderUnpackSizes + i)); - } - } - - for (;;) - { - UInt64 type; - RINOK(ReadID(&sd, &type)); - if (type == k7zIdEnd) - { - *sd2 = sd; - return SZ_OK; - } - if (type == k7zIdCRC) - { - RINOK(ReadBitUi32s(&sd, numFolders, &p->FolderCRCs, alloc)); - continue; - } - RINOK(SkipData(&sd)); - } -} - - -UInt64 SzAr_GetFolderUnpackSize(const CSzAr *p, UInt32 folderIndex) -{ - return p->CoderUnpackSizes[p->FoToCoderUnpackSizes[folderIndex] + p->FoToMainUnpackSizeIndex[folderIndex]]; -} - - -typedef struct -{ - UInt32 NumTotalSubStreams; - UInt32 NumSubDigests; - CSzData sdNumSubStreams; - CSzData sdSizes; - CSzData sdCRCs; -} CSubStreamInfo; - - -static SRes ReadSubStreamsInfo(CSzAr *p, CSzData *sd, CSubStreamInfo *ssi) -{ - UInt64 type = 0; - UInt32 numSubDigests = 0; - UInt32 numFolders = p->NumFolders; - UInt32 numUnpackStreams = numFolders; - UInt32 numUnpackSizesInData = 0; - - for (;;) - { - RINOK(ReadID(sd, &type)); - if (type == k7zIdNumUnpackStream) - { - UInt32 i; - ssi->sdNumSubStreams.Data = sd->Data; - numUnpackStreams = 0; - numSubDigests = 0; - for (i = 0; i < numFolders; i++) - { - UInt32 numStreams; - RINOK(SzReadNumber32(sd, &numStreams)); - if (numUnpackStreams > numUnpackStreams + numStreams) - return SZ_ERROR_UNSUPPORTED; - numUnpackStreams += numStreams; - if (numStreams != 0) - numUnpackSizesInData += (numStreams - 1); - if (numStreams != 1 || !SzBitWithVals_Check(&p->FolderCRCs, i)) - numSubDigests += numStreams; - } - ssi->sdNumSubStreams.Size = (size_t)(sd->Data - ssi->sdNumSubStreams.Data); - continue; - } - if (type == k7zIdCRC || type == k7zIdSize || type == k7zIdEnd) - break; - RINOK(SkipData(sd)); - } - - if (!ssi->sdNumSubStreams.Data) - { - numSubDigests = numFolders; - if (p->FolderCRCs.Defs) - numSubDigests = numFolders - CountDefinedBits(p->FolderCRCs.Defs, numFolders); - } - - ssi->NumTotalSubStreams = numUnpackStreams; - ssi->NumSubDigests = numSubDigests; - - if (type == k7zIdSize) - { - ssi->sdSizes.Data = sd->Data; - RINOK(SkipNumbers(sd, numUnpackSizesInData)); - ssi->sdSizes.Size = (size_t)(sd->Data - ssi->sdSizes.Data); - RINOK(ReadID(sd, &type)); - } - - for (;;) - { - if (type == k7zIdEnd) - return SZ_OK; - if (type == k7zIdCRC) - { - ssi->sdCRCs.Data = sd->Data; - RINOK(SkipBitUi32s(sd, numSubDigests)); - ssi->sdCRCs.Size = (size_t)(sd->Data - ssi->sdCRCs.Data); - } - else - { - RINOK(SkipData(sd)); - } - RINOK(ReadID(sd, &type)); - } -} - -static SRes SzReadStreamsInfo(CSzAr *p, - CSzData *sd, - UInt32 numFoldersMax, const CBuf *tempBufs, UInt32 numTempBufs, - UInt64 *dataOffset, - CSubStreamInfo *ssi, - ISzAllocPtr alloc) -{ - UInt64 type; - - SzData_Clear(&ssi->sdSizes); - SzData_Clear(&ssi->sdCRCs); - SzData_Clear(&ssi->sdNumSubStreams); - - *dataOffset = 0; - RINOK(ReadID(sd, &type)); - if (type == k7zIdPackInfo) - { - RINOK(ReadNumber(sd, dataOffset)); - if (*dataOffset > p->RangeLimit) - return SZ_ERROR_ARCHIVE; - RINOK(ReadPackInfo(p, sd, alloc)); - if (p->PackPositions[p->NumPackStreams] > p->RangeLimit - *dataOffset) - return SZ_ERROR_ARCHIVE; - RINOK(ReadID(sd, &type)); - } - if (type == k7zIdUnpackInfo) - { - RINOK(ReadUnpackInfo(p, sd, numFoldersMax, tempBufs, numTempBufs, alloc)); - RINOK(ReadID(sd, &type)); - } - if (type == k7zIdSubStreamsInfo) - { - RINOK(ReadSubStreamsInfo(p, sd, ssi)); - RINOK(ReadID(sd, &type)); - } - else - { - ssi->NumTotalSubStreams = p->NumFolders; - // ssi->NumSubDigests = 0; - } - - return (type == k7zIdEnd ? SZ_OK : SZ_ERROR_UNSUPPORTED); -} - -static SRes SzReadAndDecodePackedStreams( - ILookInStream *inStream, - CSzData *sd, - CBuf *tempBufs, - UInt32 numFoldersMax, - UInt64 baseOffset, - CSzAr *p, - ISzAllocPtr allocTemp) -{ - UInt64 dataStartPos; - UInt32 fo; - CSubStreamInfo ssi; - - RINOK(SzReadStreamsInfo(p, sd, numFoldersMax, NULL, 0, &dataStartPos, &ssi, allocTemp)); - - dataStartPos += baseOffset; - if (p->NumFolders == 0) - return SZ_ERROR_ARCHIVE; - - for (fo = 0; fo < p->NumFolders; fo++) - Buf_Init(tempBufs + fo); - - for (fo = 0; fo < p->NumFolders; fo++) - { - CBuf *tempBuf = tempBufs + fo; - UInt64 unpackSize = SzAr_GetFolderUnpackSize(p, fo); - if ((size_t)unpackSize != unpackSize) - return SZ_ERROR_MEM; - if (!Buf_Create(tempBuf, (size_t)unpackSize, allocTemp)) - return SZ_ERROR_MEM; - } - - for (fo = 0; fo < p->NumFolders; fo++) - { - const CBuf *tempBuf = tempBufs + fo; - RINOK(LookInStream_SeekTo(inStream, dataStartPos)); - RINOK(SzAr_DecodeFolder(p, fo, inStream, dataStartPos, tempBuf->data, tempBuf->size, allocTemp)); - } - - return SZ_OK; -} - -static SRes SzReadFileNames(const Byte *data, size_t size, UInt32 numFiles, size_t *offsets) -{ - size_t pos = 0; - *offsets++ = 0; - if (numFiles == 0) - return (size == 0) ? SZ_OK : SZ_ERROR_ARCHIVE; - if (size < 2) - return SZ_ERROR_ARCHIVE; - if (data[size - 2] != 0 || data[size - 1] != 0) - return SZ_ERROR_ARCHIVE; - do - { - const Byte *p; - if (pos == size) - return SZ_ERROR_ARCHIVE; - for (p = data + pos; - #ifdef _WIN32 - *(const UInt16 *)(const void *)p != 0 - #else - p[0] != 0 || p[1] != 0 - #endif - ; p += 2); - pos = (size_t)(p - data) + 2; - *offsets++ = (pos >> 1); - } - while (--numFiles); - return (pos == size) ? SZ_OK : SZ_ERROR_ARCHIVE; -} - -static MY_NO_INLINE SRes ReadTime(CSzBitUi64s *p, UInt32 num, - CSzData *sd2, - const CBuf *tempBufs, UInt32 numTempBufs, - ISzAllocPtr alloc) -{ - CSzData sd; - UInt32 i; - CNtfsFileTime *vals; - Byte *defs; - Byte external; - - RINOK(ReadBitVector(sd2, num, &p->Defs, alloc)); - - SZ_READ_BYTE_SD(sd2, external); - if (external == 0) - sd = *sd2; - else - { - UInt32 index; - RINOK(SzReadNumber32(sd2, &index)); - if (index >= numTempBufs) - return SZ_ERROR_ARCHIVE; - sd.Data = tempBufs[index].data; - sd.Size = tempBufs[index].size; - } - - MY_ALLOC_ZE(CNtfsFileTime, p->Vals, num, alloc); - vals = p->Vals; - defs = p->Defs; - for (i = 0; i < num; i++) - if (SzBitArray_Check(defs, i)) - { - if (sd.Size < 8) - return SZ_ERROR_ARCHIVE; - vals[i].Low = GetUi32(sd.Data); - vals[i].High = GetUi32(sd.Data + 4); - SKIP_DATA2(sd, 8); - } - else - vals[i].High = vals[i].Low = 0; - - if (external == 0) - *sd2 = sd; - - return SZ_OK; -} - - -#define NUM_ADDITIONAL_STREAMS_MAX 8 - - -static SRes SzReadHeader2( - CSzArEx *p, /* allocMain */ - CSzData *sd, - ILookInStream *inStream, - CBuf *tempBufs, UInt32 *numTempBufs, - ISzAllocPtr allocMain, - ISzAllocPtr allocTemp - ) -{ - CSubStreamInfo ssi; - -{ - UInt64 type; - - SzData_Clear(&ssi.sdSizes); - SzData_Clear(&ssi.sdCRCs); - SzData_Clear(&ssi.sdNumSubStreams); - - ssi.NumSubDigests = 0; - ssi.NumTotalSubStreams = 0; - - RINOK(ReadID(sd, &type)); - - if (type == k7zIdArchiveProperties) - { - for (;;) - { - UInt64 type2; - RINOK(ReadID(sd, &type2)); - if (type2 == k7zIdEnd) - break; - RINOK(SkipData(sd)); - } - RINOK(ReadID(sd, &type)); - } - - if (type == k7zIdAdditionalStreamsInfo) - { - CSzAr tempAr; - SRes res; - - SzAr_Init(&tempAr); - tempAr.RangeLimit = p->db.RangeLimit; - - res = SzReadAndDecodePackedStreams(inStream, sd, tempBufs, NUM_ADDITIONAL_STREAMS_MAX, - p->startPosAfterHeader, &tempAr, allocTemp); - *numTempBufs = tempAr.NumFolders; - SzAr_Free(&tempAr, allocTemp); - - if (res != SZ_OK) - return res; - RINOK(ReadID(sd, &type)); - } - - if (type == k7zIdMainStreamsInfo) - { - RINOK(SzReadStreamsInfo(&p->db, sd, (UInt32)1 << 30, tempBufs, *numTempBufs, - &p->dataPos, &ssi, allocMain)); - p->dataPos += p->startPosAfterHeader; - RINOK(ReadID(sd, &type)); - } - - if (type == k7zIdEnd) - { - return SZ_OK; - } - - if (type != k7zIdFilesInfo) - return SZ_ERROR_ARCHIVE; -} - -{ - UInt32 numFiles = 0; - UInt32 numEmptyStreams = 0; - const Byte *emptyStreams = NULL; - const Byte *emptyFiles = NULL; - - RINOK(SzReadNumber32(sd, &numFiles)); - p->NumFiles = numFiles; - - for (;;) - { - UInt64 type; - UInt64 size; - RINOK(ReadID(sd, &type)); - if (type == k7zIdEnd) - break; - RINOK(ReadNumber(sd, &size)); - if (size > sd->Size) - return SZ_ERROR_ARCHIVE; - - if (type >= ((UInt32)1 << 8)) - { - SKIP_DATA(sd, size); - } - else switch ((unsigned)type) - { - case k7zIdName: - { - size_t namesSize; - const Byte *namesData; - Byte external; - - SZ_READ_BYTE(external); - if (external == 0) - { - namesSize = (size_t)size - 1; - namesData = sd->Data; - } - else - { - UInt32 index; - RINOK(SzReadNumber32(sd, &index)); - if (index >= *numTempBufs) - return SZ_ERROR_ARCHIVE; - namesData = (tempBufs)[index].data; - namesSize = (tempBufs)[index].size; - } - - if ((namesSize & 1) != 0) - return SZ_ERROR_ARCHIVE; - MY_ALLOC(size_t, p->FileNameOffsets, numFiles + 1, allocMain); - MY_ALLOC_ZE_AND_CPY(p->FileNames, namesSize, namesData, allocMain); - RINOK(SzReadFileNames(p->FileNames, namesSize, numFiles, p->FileNameOffsets)) - if (external == 0) - { - SKIP_DATA(sd, namesSize); - } - break; - } - case k7zIdEmptyStream: - { - RINOK(RememberBitVector(sd, numFiles, &emptyStreams)); - numEmptyStreams = CountDefinedBits(emptyStreams, numFiles); - emptyFiles = NULL; - break; - } - case k7zIdEmptyFile: - { - RINOK(RememberBitVector(sd, numEmptyStreams, &emptyFiles)); - break; - } - case k7zIdWinAttrib: - { - Byte external; - CSzData sdSwitch; - CSzData *sdPtr; - SzBitUi32s_Free(&p->Attribs, allocMain); - RINOK(ReadBitVector(sd, numFiles, &p->Attribs.Defs, allocMain)); - - SZ_READ_BYTE(external); - if (external == 0) - sdPtr = sd; - else - { - UInt32 index; - RINOK(SzReadNumber32(sd, &index)); - if (index >= *numTempBufs) - return SZ_ERROR_ARCHIVE; - sdSwitch.Data = (tempBufs)[index].data; - sdSwitch.Size = (tempBufs)[index].size; - sdPtr = &sdSwitch; - } - RINOK(ReadUi32s(sdPtr, numFiles, &p->Attribs, allocMain)); - break; - } - /* - case k7zParent: - { - SzBitUi32s_Free(&p->Parents, allocMain); - RINOK(ReadBitVector(sd, numFiles, &p->Parents.Defs, allocMain)); - RINOK(SzReadSwitch(sd)); - RINOK(ReadUi32s(sd, numFiles, &p->Parents, allocMain)); - break; - } - */ - case k7zIdMTime: RINOK(ReadTime(&p->MTime, numFiles, sd, tempBufs, *numTempBufs, allocMain)); break; - case k7zIdCTime: RINOK(ReadTime(&p->CTime, numFiles, sd, tempBufs, *numTempBufs, allocMain)); break; - default: - { - SKIP_DATA(sd, size); - } - } - } - - if (numFiles - numEmptyStreams != ssi.NumTotalSubStreams) - return SZ_ERROR_ARCHIVE; - - for (;;) - { - UInt64 type; - RINOK(ReadID(sd, &type)); - if (type == k7zIdEnd) - break; - RINOK(SkipData(sd)); - } - - { - UInt32 i; - UInt32 emptyFileIndex = 0; - UInt32 folderIndex = 0; - UInt32 remSubStreams = 0; - UInt32 numSubStreams = 0; - UInt64 unpackPos = 0; - const Byte *digestsDefs = NULL; - const Byte *digestsVals = NULL; - UInt32 digestsValsIndex = 0; - UInt32 digestIndex; - Byte allDigestsDefined = 0; - Byte isDirMask = 0; - Byte crcMask = 0; - Byte mask = 0x80; - - MY_ALLOC(UInt32, p->FolderToFile, p->db.NumFolders + 1, allocMain); - MY_ALLOC_ZE(UInt32, p->FileToFolder, p->NumFiles, allocMain); - MY_ALLOC(UInt64, p->UnpackPositions, p->NumFiles + 1, allocMain); - MY_ALLOC_ZE(Byte, p->IsDirs, (p->NumFiles + 7) >> 3, allocMain); - - RINOK(SzBitUi32s_Alloc(&p->CRCs, p->NumFiles, allocMain)); - - if (ssi.sdCRCs.Size != 0) - { - SZ_READ_BYTE_SD(&ssi.sdCRCs, allDigestsDefined); - if (allDigestsDefined) - digestsVals = ssi.sdCRCs.Data; - else - { - size_t numBytes = (ssi.NumSubDigests + 7) >> 3; - digestsDefs = ssi.sdCRCs.Data; - digestsVals = digestsDefs + numBytes; - } - } - - digestIndex = 0; - - for (i = 0; i < numFiles; i++, mask >>= 1) - { - if (mask == 0) - { - UInt32 byteIndex = (i - 1) >> 3; - p->IsDirs[byteIndex] = isDirMask; - p->CRCs.Defs[byteIndex] = crcMask; - isDirMask = 0; - crcMask = 0; - mask = 0x80; - } - - p->UnpackPositions[i] = unpackPos; - p->CRCs.Vals[i] = 0; - - if (emptyStreams && SzBitArray_Check(emptyStreams, i)) - { - if (emptyFiles) - { - if (!SzBitArray_Check(emptyFiles, emptyFileIndex)) - isDirMask |= mask; - emptyFileIndex++; - } - else - isDirMask |= mask; - if (remSubStreams == 0) - { - p->FileToFolder[i] = (UInt32)-1; - continue; - } - } - - if (remSubStreams == 0) - { - for (;;) - { - if (folderIndex >= p->db.NumFolders) - return SZ_ERROR_ARCHIVE; - p->FolderToFile[folderIndex] = i; - numSubStreams = 1; - if (ssi.sdNumSubStreams.Data) - { - RINOK(SzReadNumber32(&ssi.sdNumSubStreams, &numSubStreams)); - } - remSubStreams = numSubStreams; - if (numSubStreams != 0) - break; - { - UInt64 folderUnpackSize = SzAr_GetFolderUnpackSize(&p->db, folderIndex); - unpackPos += folderUnpackSize; - if (unpackPos < folderUnpackSize) - return SZ_ERROR_ARCHIVE; - } - - folderIndex++; - } - } - - p->FileToFolder[i] = folderIndex; - - if (emptyStreams && SzBitArray_Check(emptyStreams, i)) - continue; - - if (--remSubStreams == 0) - { - UInt64 folderUnpackSize = SzAr_GetFolderUnpackSize(&p->db, folderIndex); - UInt64 startFolderUnpackPos = p->UnpackPositions[p->FolderToFile[folderIndex]]; - if (folderUnpackSize < unpackPos - startFolderUnpackPos) - return SZ_ERROR_ARCHIVE; - unpackPos = startFolderUnpackPos + folderUnpackSize; - if (unpackPos < folderUnpackSize) - return SZ_ERROR_ARCHIVE; - - if (numSubStreams == 1 && SzBitWithVals_Check(&p->db.FolderCRCs, i)) - { - p->CRCs.Vals[i] = p->db.FolderCRCs.Vals[folderIndex]; - crcMask |= mask; - } - else if (allDigestsDefined || (digestsDefs && SzBitArray_Check(digestsDefs, digestIndex))) - { - p->CRCs.Vals[i] = GetUi32(digestsVals + (size_t)digestsValsIndex * 4); - digestsValsIndex++; - crcMask |= mask; - } - - folderIndex++; - } - else - { - UInt64 v; - RINOK(ReadNumber(&ssi.sdSizes, &v)); - unpackPos += v; - if (unpackPos < v) - return SZ_ERROR_ARCHIVE; - if (allDigestsDefined || (digestsDefs && SzBitArray_Check(digestsDefs, digestIndex))) - { - p->CRCs.Vals[i] = GetUi32(digestsVals + (size_t)digestsValsIndex * 4); - digestsValsIndex++; - crcMask |= mask; - } - } - } - - if (mask != 0x80) - { - UInt32 byteIndex = (i - 1) >> 3; - p->IsDirs[byteIndex] = isDirMask; - p->CRCs.Defs[byteIndex] = crcMask; - } - - p->UnpackPositions[i] = unpackPos; - - if (remSubStreams != 0) - return SZ_ERROR_ARCHIVE; - - for (;;) - { - p->FolderToFile[folderIndex] = i; - if (folderIndex >= p->db.NumFolders) - break; - if (!ssi.sdNumSubStreams.Data) - return SZ_ERROR_ARCHIVE; - RINOK(SzReadNumber32(&ssi.sdNumSubStreams, &numSubStreams)); - if (numSubStreams != 0) - return SZ_ERROR_ARCHIVE; - /* - { - UInt64 folderUnpackSize = SzAr_GetFolderUnpackSize(&p->db, folderIndex); - unpackPos += folderUnpackSize; - if (unpackPos < folderUnpackSize) - return SZ_ERROR_ARCHIVE; - } - */ - folderIndex++; - } - - if (ssi.sdNumSubStreams.Data && ssi.sdNumSubStreams.Size != 0) - return SZ_ERROR_ARCHIVE; - } -} - return SZ_OK; -} - - -static SRes SzReadHeader( - CSzArEx *p, - CSzData *sd, - ILookInStream *inStream, - ISzAllocPtr allocMain, - ISzAllocPtr allocTemp) -{ - UInt32 i; - UInt32 numTempBufs = 0; - SRes res; - CBuf tempBufs[NUM_ADDITIONAL_STREAMS_MAX]; - - for (i = 0; i < NUM_ADDITIONAL_STREAMS_MAX; i++) - Buf_Init(tempBufs + i); - - res = SzReadHeader2(p, sd, inStream, - tempBufs, &numTempBufs, - allocMain, allocTemp); - - for (i = 0; i < NUM_ADDITIONAL_STREAMS_MAX; i++) - Buf_Free(tempBufs + i, allocTemp); - - RINOK(res); - - if (sd->Size != 0) - return SZ_ERROR_FAIL; - - return res; -} - -static SRes SzArEx_Open2( - CSzArEx *p, - ILookInStream *inStream, - ISzAllocPtr allocMain, - ISzAllocPtr allocTemp) -{ - Byte header[k7zStartHeaderSize]; - Int64 startArcPos; - UInt64 nextHeaderOffset, nextHeaderSize; - size_t nextHeaderSizeT; - UInt32 nextHeaderCRC; - CBuf buf; - SRes res; - - startArcPos = 0; - RINOK(ILookInStream_Seek(inStream, &startArcPos, SZ_SEEK_CUR)); - - RINOK(LookInStream_Read2(inStream, header, k7zStartHeaderSize, SZ_ERROR_NO_ARCHIVE)); - - if (!TestSignatureCandidate(header)) - return SZ_ERROR_NO_ARCHIVE; - if (header[6] != k7zMajorVersion) - return SZ_ERROR_UNSUPPORTED; - - nextHeaderOffset = GetUi64(header + 12); - nextHeaderSize = GetUi64(header + 20); - nextHeaderCRC = GetUi32(header + 28); - - p->startPosAfterHeader = (UInt64)startArcPos + k7zStartHeaderSize; - - if (CrcCalc(header + 12, 20) != GetUi32(header + 8)) - return SZ_ERROR_CRC; - - p->db.RangeLimit = nextHeaderOffset; - - nextHeaderSizeT = (size_t)nextHeaderSize; - if (nextHeaderSizeT != nextHeaderSize) - return SZ_ERROR_MEM; - if (nextHeaderSizeT == 0) - return SZ_OK; - if (nextHeaderOffset > nextHeaderOffset + nextHeaderSize || - nextHeaderOffset > nextHeaderOffset + nextHeaderSize + k7zStartHeaderSize) - return SZ_ERROR_NO_ARCHIVE; - - { - Int64 pos = 0; - RINOK(ILookInStream_Seek(inStream, &pos, SZ_SEEK_END)); - if ((UInt64)pos < (UInt64)startArcPos + nextHeaderOffset || - (UInt64)pos < (UInt64)startArcPos + k7zStartHeaderSize + nextHeaderOffset || - (UInt64)pos < (UInt64)startArcPos + k7zStartHeaderSize + nextHeaderOffset + nextHeaderSize) - return SZ_ERROR_INPUT_EOF; - } - - RINOK(LookInStream_SeekTo(inStream, (UInt64)startArcPos + k7zStartHeaderSize + nextHeaderOffset)); - - if (!Buf_Create(&buf, nextHeaderSizeT, allocTemp)) - return SZ_ERROR_MEM; - - res = LookInStream_Read(inStream, buf.data, nextHeaderSizeT); - - if (res == SZ_OK) - { - res = SZ_ERROR_ARCHIVE; - if (CrcCalc(buf.data, nextHeaderSizeT) == nextHeaderCRC) - { - CSzData sd; - UInt64 type; - sd.Data = buf.data; - sd.Size = buf.size; - - res = ReadID(&sd, &type); - - if (res == SZ_OK && type == k7zIdEncodedHeader) - { - CSzAr tempAr; - CBuf tempBuf; - Buf_Init(&tempBuf); - - SzAr_Init(&tempAr); - tempAr.RangeLimit = p->db.RangeLimit; - - res = SzReadAndDecodePackedStreams(inStream, &sd, &tempBuf, 1, p->startPosAfterHeader, &tempAr, allocTemp); - SzAr_Free(&tempAr, allocTemp); - - if (res != SZ_OK) - { - Buf_Free(&tempBuf, allocTemp); - } - else - { - Buf_Free(&buf, allocTemp); - buf.data = tempBuf.data; - buf.size = tempBuf.size; - sd.Data = buf.data; - sd.Size = buf.size; - res = ReadID(&sd, &type); - } - } - - if (res == SZ_OK) - { - if (type == k7zIdHeader) - { - /* - CSzData sd2; - unsigned ttt; - for (ttt = 0; ttt < 40000; ttt++) - { - SzArEx_Free(p, allocMain); - sd2 = sd; - res = SzReadHeader(p, &sd2, inStream, allocMain, allocTemp); - if (res != SZ_OK) - break; - } - */ - res = SzReadHeader(p, &sd, inStream, allocMain, allocTemp); - } - else - res = SZ_ERROR_UNSUPPORTED; - } - } - } - - Buf_Free(&buf, allocTemp); - return res; -} - - -SRes SzArEx_Open(CSzArEx *p, ILookInStream *inStream, - ISzAllocPtr allocMain, ISzAllocPtr allocTemp) -{ - SRes res = SzArEx_Open2(p, inStream, allocMain, allocTemp); - if (res != SZ_OK) - SzArEx_Free(p, allocMain); - return res; -} - - -SRes SzArEx_Extract( - const CSzArEx *p, - ILookInStream *inStream, - UInt32 fileIndex, - UInt32 *blockIndex, - Byte **tempBuf, - size_t *outBufferSize, - size_t *offset, - size_t *outSizeProcessed, - ISzAllocPtr allocMain, - ISzAllocPtr allocTemp) -{ - UInt32 folderIndex = p->FileToFolder[fileIndex]; - SRes res = SZ_OK; - - *offset = 0; - *outSizeProcessed = 0; - - if (folderIndex == (UInt32)-1) - { - ISzAlloc_Free(allocMain, *tempBuf); - *blockIndex = folderIndex; - *tempBuf = NULL; - *outBufferSize = 0; - return SZ_OK; - } - - if (*tempBuf == NULL || *blockIndex != folderIndex) - { - UInt64 unpackSizeSpec = SzAr_GetFolderUnpackSize(&p->db, folderIndex); - /* - UInt64 unpackSizeSpec = - p->UnpackPositions[p->FolderToFile[(size_t)folderIndex + 1]] - - p->UnpackPositions[p->FolderToFile[folderIndex]]; - */ - size_t unpackSize = (size_t)unpackSizeSpec; - - if (unpackSize != unpackSizeSpec) - return SZ_ERROR_MEM; - *blockIndex = folderIndex; - ISzAlloc_Free(allocMain, *tempBuf); - *tempBuf = NULL; - - if (res == SZ_OK) - { - *outBufferSize = unpackSize; - if (unpackSize != 0) - { - *tempBuf = (Byte *)ISzAlloc_Alloc(allocMain, unpackSize); - if (*tempBuf == NULL) - res = SZ_ERROR_MEM; - } - - if (res == SZ_OK) - { - res = SzAr_DecodeFolder(&p->db, folderIndex, - inStream, p->dataPos, *tempBuf, unpackSize, allocTemp); - } - } - } - - if (res == SZ_OK) - { - UInt64 unpackPos = p->UnpackPositions[fileIndex]; - *offset = (size_t)(unpackPos - p->UnpackPositions[p->FolderToFile[folderIndex]]); - *outSizeProcessed = (size_t)(p->UnpackPositions[(size_t)fileIndex + 1] - unpackPos); - if (*offset + *outSizeProcessed > *outBufferSize) - return SZ_ERROR_FAIL; - if (SzBitWithVals_Check(&p->CRCs, fileIndex)) - if (CrcCalc(*tempBuf + *offset, *outSizeProcessed) != p->CRCs.Vals[fileIndex]) - res = SZ_ERROR_CRC; - } - - return res; -} - - -size_t SzArEx_GetFileNameUtf16(const CSzArEx *p, size_t fileIndex, UInt16 *dest) -{ - size_t offs = p->FileNameOffsets[fileIndex]; - size_t len = p->FileNameOffsets[fileIndex + 1] - offs; - if (dest != 0) - { - size_t i; - const Byte *src = p->FileNames + offs * 2; - for (i = 0; i < len; i++) - dest[i] = GetUi16(src + i * 2); - } - return len; -} - -/* -size_t SzArEx_GetFullNameLen(const CSzArEx *p, size_t fileIndex) -{ - size_t len; - if (!p->FileNameOffsets) - return 1; - len = 0; - for (;;) - { - UInt32 parent = (UInt32)(Int32)-1; - len += p->FileNameOffsets[fileIndex + 1] - p->FileNameOffsets[fileIndex]; - if SzBitWithVals_Check(&p->Parents, fileIndex) - parent = p->Parents.Vals[fileIndex]; - if (parent == (UInt32)(Int32)-1) - return len; - fileIndex = parent; - } -} - -UInt16 *SzArEx_GetFullNameUtf16_Back(const CSzArEx *p, size_t fileIndex, UInt16 *dest) -{ - BoolInt needSlash; - if (!p->FileNameOffsets) - { - *(--dest) = 0; - return dest; - } - needSlash = False; - for (;;) - { - UInt32 parent = (UInt32)(Int32)-1; - size_t curLen = p->FileNameOffsets[fileIndex + 1] - p->FileNameOffsets[fileIndex]; - SzArEx_GetFileNameUtf16(p, fileIndex, dest - curLen); - if (needSlash) - *(dest - 1) = '/'; - needSlash = True; - dest -= curLen; - - if SzBitWithVals_Check(&p->Parents, fileIndex) - parent = p->Parents.Vals[fileIndex]; - if (parent == (UInt32)(Int32)-1) - return dest; - fileIndex = parent; - } -} -*/ +/* 7zArcIn.c -- 7z Input functions +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#include "7z.h" +#include "7zBuf.h" +#include "7zCrc.h" +#include "CpuArch.h" + +#define MY_ALLOC(T, p, size, alloc) { \ + if ((p = (T *)ISzAlloc_Alloc(alloc, (size) * sizeof(T))) == NULL) return SZ_ERROR_MEM; } + +#define MY_ALLOC_ZE(T, p, size, alloc) { if ((size) == 0) p = NULL; else MY_ALLOC(T, p, size, alloc) } + +#define MY_ALLOC_AND_CPY(to, size, from, alloc) \ + { MY_ALLOC(Byte, to, size, alloc); memcpy(to, from, size); } + +#define MY_ALLOC_ZE_AND_CPY(to, size, from, alloc) \ + { if ((size) == 0) to = NULL; else { MY_ALLOC_AND_CPY(to, size, from, alloc) } } + +#define k7zMajorVersion 0 + +enum EIdEnum +{ + k7zIdEnd, + k7zIdHeader, + k7zIdArchiveProperties, + k7zIdAdditionalStreamsInfo, + k7zIdMainStreamsInfo, + k7zIdFilesInfo, + k7zIdPackInfo, + k7zIdUnpackInfo, + k7zIdSubStreamsInfo, + k7zIdSize, + k7zIdCRC, + k7zIdFolder, + k7zIdCodersUnpackSize, + k7zIdNumUnpackStream, + k7zIdEmptyStream, + k7zIdEmptyFile, + k7zIdAnti, + k7zIdName, + k7zIdCTime, + k7zIdATime, + k7zIdMTime, + k7zIdWinAttrib, + k7zIdComment, + k7zIdEncodedHeader, + k7zIdStartPos, + k7zIdDummy + // k7zNtSecure, + // k7zParent, + // k7zIsReal +}; + +const Byte k7zSignature[k7zSignatureSize] = {'7', 'z', 0xBC, 0xAF, 0x27, 0x1C}; + +#define SzBitUi32s_Init(p) { (p)->Defs = NULL; (p)->Vals = NULL; } + +static SRes SzBitUi32s_Alloc(CSzBitUi32s *p, size_t num, ISzAllocPtr alloc) +{ + if (num == 0) + { + p->Defs = NULL; + p->Vals = NULL; + } + else + { + MY_ALLOC(Byte, p->Defs, (num + 7) >> 3, alloc); + MY_ALLOC(UInt32, p->Vals, num, alloc); + } + return SZ_OK; +} + +static void SzBitUi32s_Free(CSzBitUi32s *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->Defs); p->Defs = NULL; + ISzAlloc_Free(alloc, p->Vals); p->Vals = NULL; +} + +#define SzBitUi64s_Init(p) { (p)->Defs = NULL; (p)->Vals = NULL; } + +static void SzBitUi64s_Free(CSzBitUi64s *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->Defs); p->Defs = NULL; + ISzAlloc_Free(alloc, p->Vals); p->Vals = NULL; +} + + +static void SzAr_Init(CSzAr *p) +{ + p->NumPackStreams = 0; + p->NumFolders = 0; + + p->PackPositions = NULL; + SzBitUi32s_Init(&p->FolderCRCs); + + p->FoCodersOffsets = NULL; + p->FoStartPackStreamIndex = NULL; + p->FoToCoderUnpackSizes = NULL; + p->FoToMainUnpackSizeIndex = NULL; + p->CoderUnpackSizes = NULL; + + p->CodersData = NULL; + + p->RangeLimit = 0; +} + +static void SzAr_Free(CSzAr *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->PackPositions); + SzBitUi32s_Free(&p->FolderCRCs, alloc); + + ISzAlloc_Free(alloc, p->FoCodersOffsets); + ISzAlloc_Free(alloc, p->FoStartPackStreamIndex); + ISzAlloc_Free(alloc, p->FoToCoderUnpackSizes); + ISzAlloc_Free(alloc, p->FoToMainUnpackSizeIndex); + ISzAlloc_Free(alloc, p->CoderUnpackSizes); + + ISzAlloc_Free(alloc, p->CodersData); + + SzAr_Init(p); +} + + +void SzArEx_Init(CSzArEx *p) +{ + SzAr_Init(&p->db); + + p->NumFiles = 0; + p->dataPos = 0; + + p->UnpackPositions = NULL; + p->IsDirs = NULL; + + p->FolderToFile = NULL; + p->FileToFolder = NULL; + + p->FileNameOffsets = NULL; + p->FileNames = NULL; + + SzBitUi32s_Init(&p->CRCs); + SzBitUi32s_Init(&p->Attribs); + // SzBitUi32s_Init(&p->Parents); + SzBitUi64s_Init(&p->MTime); + SzBitUi64s_Init(&p->CTime); +} + +void SzArEx_Free(CSzArEx *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->UnpackPositions); + ISzAlloc_Free(alloc, p->IsDirs); + + ISzAlloc_Free(alloc, p->FolderToFile); + ISzAlloc_Free(alloc, p->FileToFolder); + + ISzAlloc_Free(alloc, p->FileNameOffsets); + ISzAlloc_Free(alloc, p->FileNames); + + SzBitUi32s_Free(&p->CRCs, alloc); + SzBitUi32s_Free(&p->Attribs, alloc); + // SzBitUi32s_Free(&p->Parents, alloc); + SzBitUi64s_Free(&p->MTime, alloc); + SzBitUi64s_Free(&p->CTime, alloc); + + SzAr_Free(&p->db, alloc); + SzArEx_Init(p); +} + + +static int TestSignatureCandidate(const Byte *testBytes) +{ + unsigned i; + for (i = 0; i < k7zSignatureSize; i++) + if (testBytes[i] != k7zSignature[i]) + return 0; + return 1; +} + +#define SzData_Clear(p) { (p)->Data = NULL; (p)->Size = 0; } + +#define SZ_READ_BYTE_SD(_sd_, dest) if ((_sd_)->Size == 0) return SZ_ERROR_ARCHIVE; (_sd_)->Size--; dest = *(_sd_)->Data++; +#define SZ_READ_BYTE(dest) SZ_READ_BYTE_SD(sd, dest) +#define SZ_READ_BYTE_2(dest) if (sd.Size == 0) return SZ_ERROR_ARCHIVE; sd.Size--; dest = *sd.Data++; + +#define SKIP_DATA(sd, size) { sd->Size -= (size_t)(size); sd->Data += (size_t)(size); } +#define SKIP_DATA2(sd, size) { sd.Size -= (size_t)(size); sd.Data += (size_t)(size); } + +#define SZ_READ_32(dest) if (sd.Size < 4) return SZ_ERROR_ARCHIVE; \ + dest = GetUi32(sd.Data); SKIP_DATA2(sd, 4); + +static MY_NO_INLINE SRes ReadNumber(CSzData *sd, UInt64 *value) +{ + Byte firstByte, mask; + unsigned i; + UInt32 v; + + SZ_READ_BYTE(firstByte); + if ((firstByte & 0x80) == 0) + { + *value = firstByte; + return SZ_OK; + } + SZ_READ_BYTE(v); + if ((firstByte & 0x40) == 0) + { + *value = (((UInt32)firstByte & 0x3F) << 8) | v; + return SZ_OK; + } + SZ_READ_BYTE(mask); + *value = v | ((UInt32)mask << 8); + mask = 0x20; + for (i = 2; i < 8; i++) + { + Byte b; + if ((firstByte & mask) == 0) + { + UInt64 highPart = (unsigned)firstByte & (unsigned)(mask - 1); + *value |= (highPart << (8 * i)); + return SZ_OK; + } + SZ_READ_BYTE(b); + *value |= ((UInt64)b << (8 * i)); + mask >>= 1; + } + return SZ_OK; +} + + +static MY_NO_INLINE SRes SzReadNumber32(CSzData *sd, UInt32 *value) +{ + Byte firstByte; + UInt64 value64; + if (sd->Size == 0) + return SZ_ERROR_ARCHIVE; + firstByte = *sd->Data; + if ((firstByte & 0x80) == 0) + { + *value = firstByte; + sd->Data++; + sd->Size--; + return SZ_OK; + } + RINOK(ReadNumber(sd, &value64)); + if (value64 >= (UInt32)0x80000000 - 1) + return SZ_ERROR_UNSUPPORTED; + if (value64 >= ((UInt64)(1) << ((sizeof(size_t) - 1) * 8 + 4))) + return SZ_ERROR_UNSUPPORTED; + *value = (UInt32)value64; + return SZ_OK; +} + +#define ReadID(sd, value) ReadNumber(sd, value) + +static SRes SkipData(CSzData *sd) +{ + UInt64 size; + RINOK(ReadNumber(sd, &size)); + if (size > sd->Size) + return SZ_ERROR_ARCHIVE; + SKIP_DATA(sd, size); + return SZ_OK; +} + +static SRes WaitId(CSzData *sd, UInt32 id) +{ + for (;;) + { + UInt64 type; + RINOK(ReadID(sd, &type)); + if (type == id) + return SZ_OK; + if (type == k7zIdEnd) + return SZ_ERROR_ARCHIVE; + RINOK(SkipData(sd)); + } +} + +static SRes RememberBitVector(CSzData *sd, UInt32 numItems, const Byte **v) +{ + UInt32 numBytes = (numItems + 7) >> 3; + if (numBytes > sd->Size) + return SZ_ERROR_ARCHIVE; + *v = sd->Data; + SKIP_DATA(sd, numBytes); + return SZ_OK; +} + +static UInt32 CountDefinedBits(const Byte *bits, UInt32 numItems) +{ + Byte b = 0; + unsigned m = 0; + UInt32 sum = 0; + for (; numItems != 0; numItems--) + { + if (m == 0) + { + b = *bits++; + m = 8; + } + m--; + sum += ((b >> m) & 1); + } + return sum; +} + +static MY_NO_INLINE SRes ReadBitVector(CSzData *sd, UInt32 numItems, Byte **v, ISzAllocPtr alloc) +{ + Byte allAreDefined; + Byte *v2; + UInt32 numBytes = (numItems + 7) >> 3; + *v = NULL; + SZ_READ_BYTE(allAreDefined); + if (numBytes == 0) + return SZ_OK; + if (allAreDefined == 0) + { + if (numBytes > sd->Size) + return SZ_ERROR_ARCHIVE; + MY_ALLOC_AND_CPY(*v, numBytes, sd->Data, alloc); + SKIP_DATA(sd, numBytes); + return SZ_OK; + } + MY_ALLOC(Byte, *v, numBytes, alloc); + v2 = *v; + memset(v2, 0xFF, (size_t)numBytes); + { + unsigned numBits = (unsigned)numItems & 7; + if (numBits != 0) + v2[(size_t)numBytes - 1] = (Byte)((((UInt32)1 << numBits) - 1) << (8 - numBits)); + } + return SZ_OK; +} + +static MY_NO_INLINE SRes ReadUi32s(CSzData *sd2, UInt32 numItems, CSzBitUi32s *crcs, ISzAllocPtr alloc) +{ + UInt32 i; + CSzData sd; + UInt32 *vals; + const Byte *defs; + MY_ALLOC_ZE(UInt32, crcs->Vals, numItems, alloc); + sd = *sd2; + defs = crcs->Defs; + vals = crcs->Vals; + for (i = 0; i < numItems; i++) + if (SzBitArray_Check(defs, i)) + { + SZ_READ_32(vals[i]); + } + else + vals[i] = 0; + *sd2 = sd; + return SZ_OK; +} + +static SRes ReadBitUi32s(CSzData *sd, UInt32 numItems, CSzBitUi32s *crcs, ISzAllocPtr alloc) +{ + SzBitUi32s_Free(crcs, alloc); + RINOK(ReadBitVector(sd, numItems, &crcs->Defs, alloc)); + return ReadUi32s(sd, numItems, crcs, alloc); +} + +static SRes SkipBitUi32s(CSzData *sd, UInt32 numItems) +{ + Byte allAreDefined; + UInt32 numDefined = numItems; + SZ_READ_BYTE(allAreDefined); + if (!allAreDefined) + { + size_t numBytes = (numItems + 7) >> 3; + if (numBytes > sd->Size) + return SZ_ERROR_ARCHIVE; + numDefined = CountDefinedBits(sd->Data, numItems); + SKIP_DATA(sd, numBytes); + } + if (numDefined > (sd->Size >> 2)) + return SZ_ERROR_ARCHIVE; + SKIP_DATA(sd, (size_t)numDefined * 4); + return SZ_OK; +} + +static SRes ReadPackInfo(CSzAr *p, CSzData *sd, ISzAllocPtr alloc) +{ + RINOK(SzReadNumber32(sd, &p->NumPackStreams)); + + RINOK(WaitId(sd, k7zIdSize)); + MY_ALLOC(UInt64, p->PackPositions, (size_t)p->NumPackStreams + 1, alloc); + { + UInt64 sum = 0; + UInt32 i; + UInt32 numPackStreams = p->NumPackStreams; + for (i = 0; i < numPackStreams; i++) + { + UInt64 packSize; + p->PackPositions[i] = sum; + RINOK(ReadNumber(sd, &packSize)); + sum += packSize; + if (sum < packSize) + return SZ_ERROR_ARCHIVE; + } + p->PackPositions[i] = sum; + } + + for (;;) + { + UInt64 type; + RINOK(ReadID(sd, &type)); + if (type == k7zIdEnd) + return SZ_OK; + if (type == k7zIdCRC) + { + /* CRC of packed streams is unused now */ + RINOK(SkipBitUi32s(sd, p->NumPackStreams)); + continue; + } + RINOK(SkipData(sd)); + } +} + +/* +static SRes SzReadSwitch(CSzData *sd) +{ + Byte external; + RINOK(SzReadByte(sd, &external)); + return (external == 0) ? SZ_OK: SZ_ERROR_UNSUPPORTED; +} +*/ + +#define k_NumCodersStreams_in_Folder_MAX (SZ_NUM_BONDS_IN_FOLDER_MAX + SZ_NUM_PACK_STREAMS_IN_FOLDER_MAX) + +SRes SzGetNextFolderItem(CSzFolder *f, CSzData *sd) +{ + UInt32 numCoders, i; + UInt32 numInStreams = 0; + const Byte *dataStart = sd->Data; + + f->NumCoders = 0; + f->NumBonds = 0; + f->NumPackStreams = 0; + f->UnpackStream = 0; + + RINOK(SzReadNumber32(sd, &numCoders)); + if (numCoders == 0 || numCoders > SZ_NUM_CODERS_IN_FOLDER_MAX) + return SZ_ERROR_UNSUPPORTED; + + for (i = 0; i < numCoders; i++) + { + Byte mainByte; + CSzCoderInfo *coder = f->Coders + i; + unsigned idSize, j; + UInt64 id; + + SZ_READ_BYTE(mainByte); + if ((mainByte & 0xC0) != 0) + return SZ_ERROR_UNSUPPORTED; + + idSize = (unsigned)(mainByte & 0xF); + if (idSize > sizeof(id)) + return SZ_ERROR_UNSUPPORTED; + if (idSize > sd->Size) + return SZ_ERROR_ARCHIVE; + id = 0; + for (j = 0; j < idSize; j++) + { + id = ((id << 8) | *sd->Data); + sd->Data++; + sd->Size--; + } + if (id > (UInt32)0xFFFFFFFF) + return SZ_ERROR_UNSUPPORTED; + coder->MethodID = (UInt32)id; + + coder->NumStreams = 1; + coder->PropsOffset = 0; + coder->PropsSize = 0; + + if ((mainByte & 0x10) != 0) + { + UInt32 numStreams; + + RINOK(SzReadNumber32(sd, &numStreams)); + if (numStreams > k_NumCodersStreams_in_Folder_MAX) + return SZ_ERROR_UNSUPPORTED; + coder->NumStreams = (Byte)numStreams; + + RINOK(SzReadNumber32(sd, &numStreams)); + if (numStreams != 1) + return SZ_ERROR_UNSUPPORTED; + } + + numInStreams += coder->NumStreams; + + if (numInStreams > k_NumCodersStreams_in_Folder_MAX) + return SZ_ERROR_UNSUPPORTED; + + if ((mainByte & 0x20) != 0) + { + UInt32 propsSize = 0; + RINOK(SzReadNumber32(sd, &propsSize)); + if (propsSize > sd->Size) + return SZ_ERROR_ARCHIVE; + if (propsSize >= 0x80) + return SZ_ERROR_UNSUPPORTED; + coder->PropsOffset = (size_t)(sd->Data - dataStart); + coder->PropsSize = (Byte)propsSize; + sd->Data += (size_t)propsSize; + sd->Size -= (size_t)propsSize; + } + } + + /* + if (numInStreams == 1 && numCoders == 1) + { + f->NumPackStreams = 1; + f->PackStreams[0] = 0; + } + else + */ + { + Byte streamUsed[k_NumCodersStreams_in_Folder_MAX]; + UInt32 numBonds, numPackStreams; + + numBonds = numCoders - 1; + if (numInStreams < numBonds) + return SZ_ERROR_ARCHIVE; + if (numBonds > SZ_NUM_BONDS_IN_FOLDER_MAX) + return SZ_ERROR_UNSUPPORTED; + f->NumBonds = numBonds; + + numPackStreams = numInStreams - numBonds; + if (numPackStreams > SZ_NUM_PACK_STREAMS_IN_FOLDER_MAX) + return SZ_ERROR_UNSUPPORTED; + f->NumPackStreams = numPackStreams; + + for (i = 0; i < numInStreams; i++) + streamUsed[i] = False; + + if (numBonds != 0) + { + Byte coderUsed[SZ_NUM_CODERS_IN_FOLDER_MAX]; + + for (i = 0; i < numCoders; i++) + coderUsed[i] = False; + + for (i = 0; i < numBonds; i++) + { + CSzBond *bp = f->Bonds + i; + + RINOK(SzReadNumber32(sd, &bp->InIndex)); + if (bp->InIndex >= numInStreams || streamUsed[bp->InIndex]) + return SZ_ERROR_ARCHIVE; + streamUsed[bp->InIndex] = True; + + RINOK(SzReadNumber32(sd, &bp->OutIndex)); + if (bp->OutIndex >= numCoders || coderUsed[bp->OutIndex]) + return SZ_ERROR_ARCHIVE; + coderUsed[bp->OutIndex] = True; + } + + for (i = 0; i < numCoders; i++) + if (!coderUsed[i]) + { + f->UnpackStream = i; + break; + } + + if (i == numCoders) + return SZ_ERROR_ARCHIVE; + } + + if (numPackStreams == 1) + { + for (i = 0; i < numInStreams; i++) + if (!streamUsed[i]) + break; + if (i == numInStreams) + return SZ_ERROR_ARCHIVE; + f->PackStreams[0] = i; + } + else + for (i = 0; i < numPackStreams; i++) + { + UInt32 index; + RINOK(SzReadNumber32(sd, &index)); + if (index >= numInStreams || streamUsed[index]) + return SZ_ERROR_ARCHIVE; + streamUsed[index] = True; + f->PackStreams[i] = index; + } + } + + f->NumCoders = numCoders; + + return SZ_OK; +} + + +static MY_NO_INLINE SRes SkipNumbers(CSzData *sd2, UInt32 num) +{ + CSzData sd; + sd = *sd2; + for (; num != 0; num--) + { + Byte firstByte, mask; + unsigned i; + SZ_READ_BYTE_2(firstByte); + if ((firstByte & 0x80) == 0) + continue; + if ((firstByte & 0x40) == 0) + { + if (sd.Size == 0) + return SZ_ERROR_ARCHIVE; + sd.Size--; + sd.Data++; + continue; + } + mask = 0x20; + for (i = 2; i < 8 && (firstByte & mask) != 0; i++) + mask >>= 1; + if (i > sd.Size) + return SZ_ERROR_ARCHIVE; + SKIP_DATA2(sd, i); + } + *sd2 = sd; + return SZ_OK; +} + + +#define k_Scan_NumCoders_MAX 64 +#define k_Scan_NumCodersStreams_in_Folder_MAX 64 + + +static SRes ReadUnpackInfo(CSzAr *p, + CSzData *sd2, + UInt32 numFoldersMax, + const CBuf *tempBufs, UInt32 numTempBufs, + ISzAllocPtr alloc) +{ + CSzData sd; + + UInt32 fo, numFolders, numCodersOutStreams, packStreamIndex; + const Byte *startBufPtr; + Byte external; + + RINOK(WaitId(sd2, k7zIdFolder)); + + RINOK(SzReadNumber32(sd2, &numFolders)); + if (numFolders > numFoldersMax) + return SZ_ERROR_UNSUPPORTED; + p->NumFolders = numFolders; + + SZ_READ_BYTE_SD(sd2, external); + if (external == 0) + sd = *sd2; + else + { + UInt32 index; + RINOK(SzReadNumber32(sd2, &index)); + if (index >= numTempBufs) + return SZ_ERROR_ARCHIVE; + sd.Data = tempBufs[index].data; + sd.Size = tempBufs[index].size; + } + + MY_ALLOC(size_t, p->FoCodersOffsets, (size_t)numFolders + 1, alloc); + MY_ALLOC(UInt32, p->FoStartPackStreamIndex, (size_t)numFolders + 1, alloc); + MY_ALLOC(UInt32, p->FoToCoderUnpackSizes, (size_t)numFolders + 1, alloc); + MY_ALLOC_ZE(Byte, p->FoToMainUnpackSizeIndex, (size_t)numFolders, alloc); + + startBufPtr = sd.Data; + + packStreamIndex = 0; + numCodersOutStreams = 0; + + for (fo = 0; fo < numFolders; fo++) + { + UInt32 numCoders, ci, numInStreams = 0; + + p->FoCodersOffsets[fo] = (size_t)(sd.Data - startBufPtr); + + RINOK(SzReadNumber32(&sd, &numCoders)); + if (numCoders == 0 || numCoders > k_Scan_NumCoders_MAX) + return SZ_ERROR_UNSUPPORTED; + + for (ci = 0; ci < numCoders; ci++) + { + Byte mainByte; + unsigned idSize; + UInt32 coderInStreams; + + SZ_READ_BYTE_2(mainByte); + if ((mainByte & 0xC0) != 0) + return SZ_ERROR_UNSUPPORTED; + idSize = (mainByte & 0xF); + if (idSize > 8) + return SZ_ERROR_UNSUPPORTED; + if (idSize > sd.Size) + return SZ_ERROR_ARCHIVE; + SKIP_DATA2(sd, idSize); + + coderInStreams = 1; + + if ((mainByte & 0x10) != 0) + { + UInt32 coderOutStreams; + RINOK(SzReadNumber32(&sd, &coderInStreams)); + RINOK(SzReadNumber32(&sd, &coderOutStreams)); + if (coderInStreams > k_Scan_NumCodersStreams_in_Folder_MAX || coderOutStreams != 1) + return SZ_ERROR_UNSUPPORTED; + } + + numInStreams += coderInStreams; + + if ((mainByte & 0x20) != 0) + { + UInt32 propsSize; + RINOK(SzReadNumber32(&sd, &propsSize)); + if (propsSize > sd.Size) + return SZ_ERROR_ARCHIVE; + SKIP_DATA2(sd, propsSize); + } + } + + { + UInt32 indexOfMainStream = 0; + UInt32 numPackStreams = 1; + + if (numCoders != 1 || numInStreams != 1) + { + Byte streamUsed[k_Scan_NumCodersStreams_in_Folder_MAX]; + Byte coderUsed[k_Scan_NumCoders_MAX]; + + UInt32 i; + UInt32 numBonds = numCoders - 1; + if (numInStreams < numBonds) + return SZ_ERROR_ARCHIVE; + + if (numInStreams > k_Scan_NumCodersStreams_in_Folder_MAX) + return SZ_ERROR_UNSUPPORTED; + + for (i = 0; i < numInStreams; i++) + streamUsed[i] = False; + for (i = 0; i < numCoders; i++) + coderUsed[i] = False; + + for (i = 0; i < numBonds; i++) + { + UInt32 index; + + RINOK(SzReadNumber32(&sd, &index)); + if (index >= numInStreams || streamUsed[index]) + return SZ_ERROR_ARCHIVE; + streamUsed[index] = True; + + RINOK(SzReadNumber32(&sd, &index)); + if (index >= numCoders || coderUsed[index]) + return SZ_ERROR_ARCHIVE; + coderUsed[index] = True; + } + + numPackStreams = numInStreams - numBonds; + + if (numPackStreams != 1) + for (i = 0; i < numPackStreams; i++) + { + UInt32 index; + RINOK(SzReadNumber32(&sd, &index)); + if (index >= numInStreams || streamUsed[index]) + return SZ_ERROR_ARCHIVE; + streamUsed[index] = True; + } + + for (i = 0; i < numCoders; i++) + if (!coderUsed[i]) + { + indexOfMainStream = i; + break; + } + + if (i == numCoders) + return SZ_ERROR_ARCHIVE; + } + + p->FoStartPackStreamIndex[fo] = packStreamIndex; + p->FoToCoderUnpackSizes[fo] = numCodersOutStreams; + p->FoToMainUnpackSizeIndex[fo] = (Byte)indexOfMainStream; + numCodersOutStreams += numCoders; + if (numCodersOutStreams < numCoders) + return SZ_ERROR_UNSUPPORTED; + if (numPackStreams > p->NumPackStreams - packStreamIndex) + return SZ_ERROR_ARCHIVE; + packStreamIndex += numPackStreams; + } + } + + p->FoToCoderUnpackSizes[fo] = numCodersOutStreams; + + { + const size_t dataSize = (size_t)(sd.Data - startBufPtr); + p->FoStartPackStreamIndex[fo] = packStreamIndex; + p->FoCodersOffsets[fo] = dataSize; + MY_ALLOC_ZE_AND_CPY(p->CodersData, dataSize, startBufPtr, alloc); + } + + if (external != 0) + { + if (sd.Size != 0) + return SZ_ERROR_ARCHIVE; + sd = *sd2; + } + + RINOK(WaitId(&sd, k7zIdCodersUnpackSize)); + + MY_ALLOC_ZE(UInt64, p->CoderUnpackSizes, (size_t)numCodersOutStreams, alloc); + { + UInt32 i; + for (i = 0; i < numCodersOutStreams; i++) + { + RINOK(ReadNumber(&sd, p->CoderUnpackSizes + i)); + } + } + + for (;;) + { + UInt64 type; + RINOK(ReadID(&sd, &type)); + if (type == k7zIdEnd) + { + *sd2 = sd; + return SZ_OK; + } + if (type == k7zIdCRC) + { + RINOK(ReadBitUi32s(&sd, numFolders, &p->FolderCRCs, alloc)); + continue; + } + RINOK(SkipData(&sd)); + } +} + + +UInt64 SzAr_GetFolderUnpackSize(const CSzAr *p, UInt32 folderIndex) +{ + return p->CoderUnpackSizes[p->FoToCoderUnpackSizes[folderIndex] + p->FoToMainUnpackSizeIndex[folderIndex]]; +} + + +typedef struct +{ + UInt32 NumTotalSubStreams; + UInt32 NumSubDigests; + CSzData sdNumSubStreams; + CSzData sdSizes; + CSzData sdCRCs; +} CSubStreamInfo; + + +static SRes ReadSubStreamsInfo(CSzAr *p, CSzData *sd, CSubStreamInfo *ssi) +{ + UInt64 type = 0; + UInt32 numSubDigests = 0; + UInt32 numFolders = p->NumFolders; + UInt32 numUnpackStreams = numFolders; + UInt32 numUnpackSizesInData = 0; + + for (;;) + { + RINOK(ReadID(sd, &type)); + if (type == k7zIdNumUnpackStream) + { + UInt32 i; + ssi->sdNumSubStreams.Data = sd->Data; + numUnpackStreams = 0; + numSubDigests = 0; + for (i = 0; i < numFolders; i++) + { + UInt32 numStreams; + RINOK(SzReadNumber32(sd, &numStreams)); + if (numUnpackStreams > numUnpackStreams + numStreams) + return SZ_ERROR_UNSUPPORTED; + numUnpackStreams += numStreams; + if (numStreams != 0) + numUnpackSizesInData += (numStreams - 1); + if (numStreams != 1 || !SzBitWithVals_Check(&p->FolderCRCs, i)) + numSubDigests += numStreams; + } + ssi->sdNumSubStreams.Size = (size_t)(sd->Data - ssi->sdNumSubStreams.Data); + continue; + } + if (type == k7zIdCRC || type == k7zIdSize || type == k7zIdEnd) + break; + RINOK(SkipData(sd)); + } + + if (!ssi->sdNumSubStreams.Data) + { + numSubDigests = numFolders; + if (p->FolderCRCs.Defs) + numSubDigests = numFolders - CountDefinedBits(p->FolderCRCs.Defs, numFolders); + } + + ssi->NumTotalSubStreams = numUnpackStreams; + ssi->NumSubDigests = numSubDigests; + + if (type == k7zIdSize) + { + ssi->sdSizes.Data = sd->Data; + RINOK(SkipNumbers(sd, numUnpackSizesInData)); + ssi->sdSizes.Size = (size_t)(sd->Data - ssi->sdSizes.Data); + RINOK(ReadID(sd, &type)); + } + + for (;;) + { + if (type == k7zIdEnd) + return SZ_OK; + if (type == k7zIdCRC) + { + ssi->sdCRCs.Data = sd->Data; + RINOK(SkipBitUi32s(sd, numSubDigests)); + ssi->sdCRCs.Size = (size_t)(sd->Data - ssi->sdCRCs.Data); + } + else + { + RINOK(SkipData(sd)); + } + RINOK(ReadID(sd, &type)); + } +} + +static SRes SzReadStreamsInfo(CSzAr *p, + CSzData *sd, + UInt32 numFoldersMax, const CBuf *tempBufs, UInt32 numTempBufs, + UInt64 *dataOffset, + CSubStreamInfo *ssi, + ISzAllocPtr alloc) +{ + UInt64 type; + + SzData_Clear(&ssi->sdSizes); + SzData_Clear(&ssi->sdCRCs); + SzData_Clear(&ssi->sdNumSubStreams); + + *dataOffset = 0; + RINOK(ReadID(sd, &type)); + if (type == k7zIdPackInfo) + { + RINOK(ReadNumber(sd, dataOffset)); + if (*dataOffset > p->RangeLimit) + return SZ_ERROR_ARCHIVE; + RINOK(ReadPackInfo(p, sd, alloc)); + if (p->PackPositions[p->NumPackStreams] > p->RangeLimit - *dataOffset) + return SZ_ERROR_ARCHIVE; + RINOK(ReadID(sd, &type)); + } + if (type == k7zIdUnpackInfo) + { + RINOK(ReadUnpackInfo(p, sd, numFoldersMax, tempBufs, numTempBufs, alloc)); + RINOK(ReadID(sd, &type)); + } + if (type == k7zIdSubStreamsInfo) + { + RINOK(ReadSubStreamsInfo(p, sd, ssi)); + RINOK(ReadID(sd, &type)); + } + else + { + ssi->NumTotalSubStreams = p->NumFolders; + // ssi->NumSubDigests = 0; + } + + return (type == k7zIdEnd ? SZ_OK : SZ_ERROR_UNSUPPORTED); +} + +static SRes SzReadAndDecodePackedStreams( + ILookInStream *inStream, + CSzData *sd, + CBuf *tempBufs, + UInt32 numFoldersMax, + UInt64 baseOffset, + CSzAr *p, + ISzAllocPtr allocTemp) +{ + UInt64 dataStartPos; + UInt32 fo; + CSubStreamInfo ssi; + + RINOK(SzReadStreamsInfo(p, sd, numFoldersMax, NULL, 0, &dataStartPos, &ssi, allocTemp)); + + dataStartPos += baseOffset; + if (p->NumFolders == 0) + return SZ_ERROR_ARCHIVE; + + for (fo = 0; fo < p->NumFolders; fo++) + Buf_Init(tempBufs + fo); + + for (fo = 0; fo < p->NumFolders; fo++) + { + CBuf *tempBuf = tempBufs + fo; + UInt64 unpackSize = SzAr_GetFolderUnpackSize(p, fo); + if ((size_t)unpackSize != unpackSize) + return SZ_ERROR_MEM; + if (!Buf_Create(tempBuf, (size_t)unpackSize, allocTemp)) + return SZ_ERROR_MEM; + } + + for (fo = 0; fo < p->NumFolders; fo++) + { + const CBuf *tempBuf = tempBufs + fo; + RINOK(LookInStream_SeekTo(inStream, dataStartPos)); + RINOK(SzAr_DecodeFolder(p, fo, inStream, dataStartPos, tempBuf->data, tempBuf->size, allocTemp)); + } + + return SZ_OK; +} + +static SRes SzReadFileNames(const Byte *data, size_t size, UInt32 numFiles, size_t *offsets) +{ + size_t pos = 0; + *offsets++ = 0; + if (numFiles == 0) + return (size == 0) ? SZ_OK : SZ_ERROR_ARCHIVE; + if (size < 2) + return SZ_ERROR_ARCHIVE; + if (data[size - 2] != 0 || data[size - 1] != 0) + return SZ_ERROR_ARCHIVE; + do + { + const Byte *p; + if (pos == size) + return SZ_ERROR_ARCHIVE; + for (p = data + pos; + #ifdef _WIN32 + *(const UInt16 *)(const void *)p != 0 + #else + p[0] != 0 || p[1] != 0 + #endif + ; p += 2); + pos = (size_t)(p - data) + 2; + *offsets++ = (pos >> 1); + } + while (--numFiles); + return (pos == size) ? SZ_OK : SZ_ERROR_ARCHIVE; +} + +static MY_NO_INLINE SRes ReadTime(CSzBitUi64s *p, UInt32 num, + CSzData *sd2, + const CBuf *tempBufs, UInt32 numTempBufs, + ISzAllocPtr alloc) +{ + CSzData sd; + UInt32 i; + CNtfsFileTime *vals; + Byte *defs; + Byte external; + + RINOK(ReadBitVector(sd2, num, &p->Defs, alloc)); + + SZ_READ_BYTE_SD(sd2, external); + if (external == 0) + sd = *sd2; + else + { + UInt32 index; + RINOK(SzReadNumber32(sd2, &index)); + if (index >= numTempBufs) + return SZ_ERROR_ARCHIVE; + sd.Data = tempBufs[index].data; + sd.Size = tempBufs[index].size; + } + + MY_ALLOC_ZE(CNtfsFileTime, p->Vals, num, alloc); + vals = p->Vals; + defs = p->Defs; + for (i = 0; i < num; i++) + if (SzBitArray_Check(defs, i)) + { + if (sd.Size < 8) + return SZ_ERROR_ARCHIVE; + vals[i].Low = GetUi32(sd.Data); + vals[i].High = GetUi32(sd.Data + 4); + SKIP_DATA2(sd, 8); + } + else + vals[i].High = vals[i].Low = 0; + + if (external == 0) + *sd2 = sd; + + return SZ_OK; +} + + +#define NUM_ADDITIONAL_STREAMS_MAX 8 + + +static SRes SzReadHeader2( + CSzArEx *p, /* allocMain */ + CSzData *sd, + ILookInStream *inStream, + CBuf *tempBufs, UInt32 *numTempBufs, + ISzAllocPtr allocMain, + ISzAllocPtr allocTemp + ) +{ + CSubStreamInfo ssi; + +{ + UInt64 type; + + SzData_Clear(&ssi.sdSizes); + SzData_Clear(&ssi.sdCRCs); + SzData_Clear(&ssi.sdNumSubStreams); + + ssi.NumSubDigests = 0; + ssi.NumTotalSubStreams = 0; + + RINOK(ReadID(sd, &type)); + + if (type == k7zIdArchiveProperties) + { + for (;;) + { + UInt64 type2; + RINOK(ReadID(sd, &type2)); + if (type2 == k7zIdEnd) + break; + RINOK(SkipData(sd)); + } + RINOK(ReadID(sd, &type)); + } + + if (type == k7zIdAdditionalStreamsInfo) + { + CSzAr tempAr; + SRes res; + + SzAr_Init(&tempAr); + tempAr.RangeLimit = p->db.RangeLimit; + + res = SzReadAndDecodePackedStreams(inStream, sd, tempBufs, NUM_ADDITIONAL_STREAMS_MAX, + p->startPosAfterHeader, &tempAr, allocTemp); + *numTempBufs = tempAr.NumFolders; + SzAr_Free(&tempAr, allocTemp); + + if (res != SZ_OK) + return res; + RINOK(ReadID(sd, &type)); + } + + if (type == k7zIdMainStreamsInfo) + { + RINOK(SzReadStreamsInfo(&p->db, sd, (UInt32)1 << 30, tempBufs, *numTempBufs, + &p->dataPos, &ssi, allocMain)); + p->dataPos += p->startPosAfterHeader; + RINOK(ReadID(sd, &type)); + } + + if (type == k7zIdEnd) + { + return SZ_OK; + } + + if (type != k7zIdFilesInfo) + return SZ_ERROR_ARCHIVE; +} + +{ + UInt32 numFiles = 0; + UInt32 numEmptyStreams = 0; + const Byte *emptyStreams = NULL; + const Byte *emptyFiles = NULL; + + RINOK(SzReadNumber32(sd, &numFiles)); + p->NumFiles = numFiles; + + for (;;) + { + UInt64 type; + UInt64 size; + RINOK(ReadID(sd, &type)); + if (type == k7zIdEnd) + break; + RINOK(ReadNumber(sd, &size)); + if (size > sd->Size) + return SZ_ERROR_ARCHIVE; + + if (type >= ((UInt32)1 << 8)) + { + SKIP_DATA(sd, size); + } + else switch ((unsigned)type) + { + case k7zIdName: + { + size_t namesSize; + const Byte *namesData; + Byte external; + + SZ_READ_BYTE(external); + if (external == 0) + { + namesSize = (size_t)size - 1; + namesData = sd->Data; + } + else + { + UInt32 index; + RINOK(SzReadNumber32(sd, &index)); + if (index >= *numTempBufs) + return SZ_ERROR_ARCHIVE; + namesData = (tempBufs)[index].data; + namesSize = (tempBufs)[index].size; + } + + if ((namesSize & 1) != 0) + return SZ_ERROR_ARCHIVE; + MY_ALLOC(size_t, p->FileNameOffsets, numFiles + 1, allocMain); + MY_ALLOC_ZE_AND_CPY(p->FileNames, namesSize, namesData, allocMain); + RINOK(SzReadFileNames(p->FileNames, namesSize, numFiles, p->FileNameOffsets)) + if (external == 0) + { + SKIP_DATA(sd, namesSize); + } + break; + } + case k7zIdEmptyStream: + { + RINOK(RememberBitVector(sd, numFiles, &emptyStreams)); + numEmptyStreams = CountDefinedBits(emptyStreams, numFiles); + emptyFiles = NULL; + break; + } + case k7zIdEmptyFile: + { + RINOK(RememberBitVector(sd, numEmptyStreams, &emptyFiles)); + break; + } + case k7zIdWinAttrib: + { + Byte external; + CSzData sdSwitch; + CSzData *sdPtr; + SzBitUi32s_Free(&p->Attribs, allocMain); + RINOK(ReadBitVector(sd, numFiles, &p->Attribs.Defs, allocMain)); + + SZ_READ_BYTE(external); + if (external == 0) + sdPtr = sd; + else + { + UInt32 index; + RINOK(SzReadNumber32(sd, &index)); + if (index >= *numTempBufs) + return SZ_ERROR_ARCHIVE; + sdSwitch.Data = (tempBufs)[index].data; + sdSwitch.Size = (tempBufs)[index].size; + sdPtr = &sdSwitch; + } + RINOK(ReadUi32s(sdPtr, numFiles, &p->Attribs, allocMain)); + break; + } + /* + case k7zParent: + { + SzBitUi32s_Free(&p->Parents, allocMain); + RINOK(ReadBitVector(sd, numFiles, &p->Parents.Defs, allocMain)); + RINOK(SzReadSwitch(sd)); + RINOK(ReadUi32s(sd, numFiles, &p->Parents, allocMain)); + break; + } + */ + case k7zIdMTime: RINOK(ReadTime(&p->MTime, numFiles, sd, tempBufs, *numTempBufs, allocMain)); break; + case k7zIdCTime: RINOK(ReadTime(&p->CTime, numFiles, sd, tempBufs, *numTempBufs, allocMain)); break; + default: + { + SKIP_DATA(sd, size); + } + } + } + + if (numFiles - numEmptyStreams != ssi.NumTotalSubStreams) + return SZ_ERROR_ARCHIVE; + + for (;;) + { + UInt64 type; + RINOK(ReadID(sd, &type)); + if (type == k7zIdEnd) + break; + RINOK(SkipData(sd)); + } + + { + UInt32 i; + UInt32 emptyFileIndex = 0; + UInt32 folderIndex = 0; + UInt32 remSubStreams = 0; + UInt32 numSubStreams = 0; + UInt64 unpackPos = 0; + const Byte *digestsDefs = NULL; + const Byte *digestsVals = NULL; + UInt32 digestsValsIndex = 0; + UInt32 digestIndex; + Byte allDigestsDefined = 0; + Byte isDirMask = 0; + Byte crcMask = 0; + Byte mask = 0x80; + + MY_ALLOC(UInt32, p->FolderToFile, p->db.NumFolders + 1, allocMain); + MY_ALLOC_ZE(UInt32, p->FileToFolder, p->NumFiles, allocMain); + MY_ALLOC(UInt64, p->UnpackPositions, p->NumFiles + 1, allocMain); + MY_ALLOC_ZE(Byte, p->IsDirs, (p->NumFiles + 7) >> 3, allocMain); + + RINOK(SzBitUi32s_Alloc(&p->CRCs, p->NumFiles, allocMain)); + + if (ssi.sdCRCs.Size != 0) + { + SZ_READ_BYTE_SD(&ssi.sdCRCs, allDigestsDefined); + if (allDigestsDefined) + digestsVals = ssi.sdCRCs.Data; + else + { + size_t numBytes = (ssi.NumSubDigests + 7) >> 3; + digestsDefs = ssi.sdCRCs.Data; + digestsVals = digestsDefs + numBytes; + } + } + + digestIndex = 0; + + for (i = 0; i < numFiles; i++, mask >>= 1) + { + if (mask == 0) + { + UInt32 byteIndex = (i - 1) >> 3; + p->IsDirs[byteIndex] = isDirMask; + p->CRCs.Defs[byteIndex] = crcMask; + isDirMask = 0; + crcMask = 0; + mask = 0x80; + } + + p->UnpackPositions[i] = unpackPos; + p->CRCs.Vals[i] = 0; + + if (emptyStreams && SzBitArray_Check(emptyStreams, i)) + { + if (emptyFiles) + { + if (!SzBitArray_Check(emptyFiles, emptyFileIndex)) + isDirMask |= mask; + emptyFileIndex++; + } + else + isDirMask |= mask; + if (remSubStreams == 0) + { + p->FileToFolder[i] = (UInt32)-1; + continue; + } + } + + if (remSubStreams == 0) + { + for (;;) + { + if (folderIndex >= p->db.NumFolders) + return SZ_ERROR_ARCHIVE; + p->FolderToFile[folderIndex] = i; + numSubStreams = 1; + if (ssi.sdNumSubStreams.Data) + { + RINOK(SzReadNumber32(&ssi.sdNumSubStreams, &numSubStreams)); + } + remSubStreams = numSubStreams; + if (numSubStreams != 0) + break; + { + UInt64 folderUnpackSize = SzAr_GetFolderUnpackSize(&p->db, folderIndex); + unpackPos += folderUnpackSize; + if (unpackPos < folderUnpackSize) + return SZ_ERROR_ARCHIVE; + } + + folderIndex++; + } + } + + p->FileToFolder[i] = folderIndex; + + if (emptyStreams && SzBitArray_Check(emptyStreams, i)) + continue; + + if (--remSubStreams == 0) + { + UInt64 folderUnpackSize = SzAr_GetFolderUnpackSize(&p->db, folderIndex); + UInt64 startFolderUnpackPos = p->UnpackPositions[p->FolderToFile[folderIndex]]; + if (folderUnpackSize < unpackPos - startFolderUnpackPos) + return SZ_ERROR_ARCHIVE; + unpackPos = startFolderUnpackPos + folderUnpackSize; + if (unpackPos < folderUnpackSize) + return SZ_ERROR_ARCHIVE; + + if (numSubStreams == 1 && SzBitWithVals_Check(&p->db.FolderCRCs, i)) + { + p->CRCs.Vals[i] = p->db.FolderCRCs.Vals[folderIndex]; + crcMask |= mask; + } + else if (allDigestsDefined || (digestsDefs && SzBitArray_Check(digestsDefs, digestIndex))) + { + p->CRCs.Vals[i] = GetUi32(digestsVals + (size_t)digestsValsIndex * 4); + digestsValsIndex++; + crcMask |= mask; + } + + folderIndex++; + } + else + { + UInt64 v; + RINOK(ReadNumber(&ssi.sdSizes, &v)); + unpackPos += v; + if (unpackPos < v) + return SZ_ERROR_ARCHIVE; + if (allDigestsDefined || (digestsDefs && SzBitArray_Check(digestsDefs, digestIndex))) + { + p->CRCs.Vals[i] = GetUi32(digestsVals + (size_t)digestsValsIndex * 4); + digestsValsIndex++; + crcMask |= mask; + } + } + } + + if (mask != 0x80) + { + UInt32 byteIndex = (i - 1) >> 3; + p->IsDirs[byteIndex] = isDirMask; + p->CRCs.Defs[byteIndex] = crcMask; + } + + p->UnpackPositions[i] = unpackPos; + + if (remSubStreams != 0) + return SZ_ERROR_ARCHIVE; + + for (;;) + { + p->FolderToFile[folderIndex] = i; + if (folderIndex >= p->db.NumFolders) + break; + if (!ssi.sdNumSubStreams.Data) + return SZ_ERROR_ARCHIVE; + RINOK(SzReadNumber32(&ssi.sdNumSubStreams, &numSubStreams)); + if (numSubStreams != 0) + return SZ_ERROR_ARCHIVE; + /* + { + UInt64 folderUnpackSize = SzAr_GetFolderUnpackSize(&p->db, folderIndex); + unpackPos += folderUnpackSize; + if (unpackPos < folderUnpackSize) + return SZ_ERROR_ARCHIVE; + } + */ + folderIndex++; + } + + if (ssi.sdNumSubStreams.Data && ssi.sdNumSubStreams.Size != 0) + return SZ_ERROR_ARCHIVE; + } +} + return SZ_OK; +} + + +static SRes SzReadHeader( + CSzArEx *p, + CSzData *sd, + ILookInStream *inStream, + ISzAllocPtr allocMain, + ISzAllocPtr allocTemp) +{ + UInt32 i; + UInt32 numTempBufs = 0; + SRes res; + CBuf tempBufs[NUM_ADDITIONAL_STREAMS_MAX]; + + for (i = 0; i < NUM_ADDITIONAL_STREAMS_MAX; i++) + Buf_Init(tempBufs + i); + + res = SzReadHeader2(p, sd, inStream, + tempBufs, &numTempBufs, + allocMain, allocTemp); + + for (i = 0; i < NUM_ADDITIONAL_STREAMS_MAX; i++) + Buf_Free(tempBufs + i, allocTemp); + + RINOK(res); + + if (sd->Size != 0) + return SZ_ERROR_FAIL; + + return res; +} + +static SRes SzArEx_Open2( + CSzArEx *p, + ILookInStream *inStream, + ISzAllocPtr allocMain, + ISzAllocPtr allocTemp) +{ + Byte header[k7zStartHeaderSize]; + Int64 startArcPos; + UInt64 nextHeaderOffset, nextHeaderSize; + size_t nextHeaderSizeT; + UInt32 nextHeaderCRC; + CBuf buf; + SRes res; + + startArcPos = 0; + RINOK(ILookInStream_Seek(inStream, &startArcPos, SZ_SEEK_CUR)); + + RINOK(LookInStream_Read2(inStream, header, k7zStartHeaderSize, SZ_ERROR_NO_ARCHIVE)); + + if (!TestSignatureCandidate(header)) + return SZ_ERROR_NO_ARCHIVE; + if (header[6] != k7zMajorVersion) + return SZ_ERROR_UNSUPPORTED; + + nextHeaderOffset = GetUi64(header + 12); + nextHeaderSize = GetUi64(header + 20); + nextHeaderCRC = GetUi32(header + 28); + + p->startPosAfterHeader = (UInt64)startArcPos + k7zStartHeaderSize; + + if (CrcCalc(header + 12, 20) != GetUi32(header + 8)) + return SZ_ERROR_CRC; + + p->db.RangeLimit = nextHeaderOffset; + + nextHeaderSizeT = (size_t)nextHeaderSize; + if (nextHeaderSizeT != nextHeaderSize) + return SZ_ERROR_MEM; + if (nextHeaderSizeT == 0) + return SZ_OK; + if (nextHeaderOffset > nextHeaderOffset + nextHeaderSize || + nextHeaderOffset > nextHeaderOffset + nextHeaderSize + k7zStartHeaderSize) + return SZ_ERROR_NO_ARCHIVE; + + { + Int64 pos = 0; + RINOK(ILookInStream_Seek(inStream, &pos, SZ_SEEK_END)); + if ((UInt64)pos < (UInt64)startArcPos + nextHeaderOffset || + (UInt64)pos < (UInt64)startArcPos + k7zStartHeaderSize + nextHeaderOffset || + (UInt64)pos < (UInt64)startArcPos + k7zStartHeaderSize + nextHeaderOffset + nextHeaderSize) + return SZ_ERROR_INPUT_EOF; + } + + RINOK(LookInStream_SeekTo(inStream, (UInt64)startArcPos + k7zStartHeaderSize + nextHeaderOffset)); + + if (!Buf_Create(&buf, nextHeaderSizeT, allocTemp)) + return SZ_ERROR_MEM; + + res = LookInStream_Read(inStream, buf.data, nextHeaderSizeT); + + if (res == SZ_OK) + { + res = SZ_ERROR_ARCHIVE; + if (CrcCalc(buf.data, nextHeaderSizeT) == nextHeaderCRC) + { + CSzData sd; + UInt64 type; + sd.Data = buf.data; + sd.Size = buf.size; + + res = ReadID(&sd, &type); + + if (res == SZ_OK && type == k7zIdEncodedHeader) + { + CSzAr tempAr; + CBuf tempBuf; + Buf_Init(&tempBuf); + + SzAr_Init(&tempAr); + tempAr.RangeLimit = p->db.RangeLimit; + + res = SzReadAndDecodePackedStreams(inStream, &sd, &tempBuf, 1, p->startPosAfterHeader, &tempAr, allocTemp); + SzAr_Free(&tempAr, allocTemp); + + if (res != SZ_OK) + { + Buf_Free(&tempBuf, allocTemp); + } + else + { + Buf_Free(&buf, allocTemp); + buf.data = tempBuf.data; + buf.size = tempBuf.size; + sd.Data = buf.data; + sd.Size = buf.size; + res = ReadID(&sd, &type); + } + } + + if (res == SZ_OK) + { + if (type == k7zIdHeader) + { + /* + CSzData sd2; + unsigned ttt; + for (ttt = 0; ttt < 40000; ttt++) + { + SzArEx_Free(p, allocMain); + sd2 = sd; + res = SzReadHeader(p, &sd2, inStream, allocMain, allocTemp); + if (res != SZ_OK) + break; + } + */ + res = SzReadHeader(p, &sd, inStream, allocMain, allocTemp); + } + else + res = SZ_ERROR_UNSUPPORTED; + } + } + } + + Buf_Free(&buf, allocTemp); + return res; +} + + +SRes SzArEx_Open(CSzArEx *p, ILookInStream *inStream, + ISzAllocPtr allocMain, ISzAllocPtr allocTemp) +{ + SRes res = SzArEx_Open2(p, inStream, allocMain, allocTemp); + if (res != SZ_OK) + SzArEx_Free(p, allocMain); + return res; +} + + +SRes SzArEx_Extract( + const CSzArEx *p, + ILookInStream *inStream, + UInt32 fileIndex, + UInt32 *blockIndex, + Byte **tempBuf, + size_t *outBufferSize, + size_t *offset, + size_t *outSizeProcessed, + ISzAllocPtr allocMain, + ISzAllocPtr allocTemp) +{ + UInt32 folderIndex = p->FileToFolder[fileIndex]; + SRes res = SZ_OK; + + *offset = 0; + *outSizeProcessed = 0; + + if (folderIndex == (UInt32)-1) + { + ISzAlloc_Free(allocMain, *tempBuf); + *blockIndex = folderIndex; + *tempBuf = NULL; + *outBufferSize = 0; + return SZ_OK; + } + + if (*tempBuf == NULL || *blockIndex != folderIndex) + { + UInt64 unpackSizeSpec = SzAr_GetFolderUnpackSize(&p->db, folderIndex); + /* + UInt64 unpackSizeSpec = + p->UnpackPositions[p->FolderToFile[(size_t)folderIndex + 1]] - + p->UnpackPositions[p->FolderToFile[folderIndex]]; + */ + size_t unpackSize = (size_t)unpackSizeSpec; + + if (unpackSize != unpackSizeSpec) + return SZ_ERROR_MEM; + *blockIndex = folderIndex; + ISzAlloc_Free(allocMain, *tempBuf); + *tempBuf = NULL; + + if (res == SZ_OK) + { + *outBufferSize = unpackSize; + if (unpackSize != 0) + { + *tempBuf = (Byte *)ISzAlloc_Alloc(allocMain, unpackSize); + if (*tempBuf == NULL) + res = SZ_ERROR_MEM; + } + + if (res == SZ_OK) + { + res = SzAr_DecodeFolder(&p->db, folderIndex, + inStream, p->dataPos, *tempBuf, unpackSize, allocTemp); + } + } + } + + if (res == SZ_OK) + { + UInt64 unpackPos = p->UnpackPositions[fileIndex]; + *offset = (size_t)(unpackPos - p->UnpackPositions[p->FolderToFile[folderIndex]]); + *outSizeProcessed = (size_t)(p->UnpackPositions[(size_t)fileIndex + 1] - unpackPos); + if (*offset + *outSizeProcessed > *outBufferSize) + return SZ_ERROR_FAIL; + if (SzBitWithVals_Check(&p->CRCs, fileIndex)) + if (CrcCalc(*tempBuf + *offset, *outSizeProcessed) != p->CRCs.Vals[fileIndex]) + res = SZ_ERROR_CRC; + } + + return res; +} + + +size_t SzArEx_GetFileNameUtf16(const CSzArEx *p, size_t fileIndex, UInt16 *dest) +{ + size_t offs = p->FileNameOffsets[fileIndex]; + size_t len = p->FileNameOffsets[fileIndex + 1] - offs; + if (dest != 0) + { + size_t i; + const Byte *src = p->FileNames + offs * 2; + for (i = 0; i < len; i++) + dest[i] = GetUi16(src + i * 2); + } + return len; +} + +/* +size_t SzArEx_GetFullNameLen(const CSzArEx *p, size_t fileIndex) +{ + size_t len; + if (!p->FileNameOffsets) + return 1; + len = 0; + for (;;) + { + UInt32 parent = (UInt32)(Int32)-1; + len += p->FileNameOffsets[fileIndex + 1] - p->FileNameOffsets[fileIndex]; + if SzBitWithVals_Check(&p->Parents, fileIndex) + parent = p->Parents.Vals[fileIndex]; + if (parent == (UInt32)(Int32)-1) + return len; + fileIndex = parent; + } +} + +UInt16 *SzArEx_GetFullNameUtf16_Back(const CSzArEx *p, size_t fileIndex, UInt16 *dest) +{ + BoolInt needSlash; + if (!p->FileNameOffsets) + { + *(--dest) = 0; + return dest; + } + needSlash = False; + for (;;) + { + UInt32 parent = (UInt32)(Int32)-1; + size_t curLen = p->FileNameOffsets[fileIndex + 1] - p->FileNameOffsets[fileIndex]; + SzArEx_GetFileNameUtf16(p, fileIndex, dest - curLen); + if (needSlash) + *(dest - 1) = '/'; + needSlash = True; + dest -= curLen; + + if SzBitWithVals_Check(&p->Parents, fileIndex) + parent = p->Parents.Vals[fileIndex]; + if (parent == (UInt32)(Int32)-1) + return dest; + fileIndex = parent; + } +} +*/ diff --git a/deps/LZMA-SDK/C/7zBuf.c b/deps/LZMA-SDK/C/7zBuf.c index 438bba68b..8865c32a8 100644 --- a/deps/LZMA-SDK/C/7zBuf.c +++ b/deps/LZMA-SDK/C/7zBuf.c @@ -1,36 +1,36 @@ -/* 7zBuf.c -- Byte Buffer -2017-04-03 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "7zBuf.h" - -void Buf_Init(CBuf *p) -{ - p->data = 0; - p->size = 0; -} - -int Buf_Create(CBuf *p, size_t size, ISzAllocPtr alloc) -{ - p->size = 0; - if (size == 0) - { - p->data = 0; - return 1; - } - p->data = (Byte *)ISzAlloc_Alloc(alloc, size); - if (p->data) - { - p->size = size; - return 1; - } - return 0; -} - -void Buf_Free(CBuf *p, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, p->data); - p->data = 0; - p->size = 0; -} +/* 7zBuf.c -- Byte Buffer +2017-04-03 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "7zBuf.h" + +void Buf_Init(CBuf *p) +{ + p->data = 0; + p->size = 0; +} + +int Buf_Create(CBuf *p, size_t size, ISzAllocPtr alloc) +{ + p->size = 0; + if (size == 0) + { + p->data = 0; + return 1; + } + p->data = (Byte *)ISzAlloc_Alloc(alloc, size); + if (p->data) + { + p->size = size; + return 1; + } + return 0; +} + +void Buf_Free(CBuf *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->data); + p->data = 0; + p->size = 0; +} diff --git a/deps/LZMA-SDK/C/7zBuf.h b/deps/LZMA-SDK/C/7zBuf.h index 5942d6e62..81d1b5b64 100644 --- a/deps/LZMA-SDK/C/7zBuf.h +++ b/deps/LZMA-SDK/C/7zBuf.h @@ -1,35 +1,35 @@ -/* 7zBuf.h -- Byte Buffer -2017-04-03 : Igor Pavlov : Public domain */ - -#ifndef __7Z_BUF_H -#define __7Z_BUF_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -typedef struct -{ - Byte *data; - size_t size; -} CBuf; - -void Buf_Init(CBuf *p); -int Buf_Create(CBuf *p, size_t size, ISzAllocPtr alloc); -void Buf_Free(CBuf *p, ISzAllocPtr alloc); - -typedef struct -{ - Byte *data; - size_t size; - size_t pos; -} CDynBuf; - -void DynBuf_Construct(CDynBuf *p); -void DynBuf_SeekToBeg(CDynBuf *p); -int DynBuf_Write(CDynBuf *p, const Byte *buf, size_t size, ISzAllocPtr alloc); -void DynBuf_Free(CDynBuf *p, ISzAllocPtr alloc); - -EXTERN_C_END - -#endif +/* 7zBuf.h -- Byte Buffer +2017-04-03 : Igor Pavlov : Public domain */ + +#ifndef __7Z_BUF_H +#define __7Z_BUF_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +typedef struct +{ + Byte *data; + size_t size; +} CBuf; + +void Buf_Init(CBuf *p); +int Buf_Create(CBuf *p, size_t size, ISzAllocPtr alloc); +void Buf_Free(CBuf *p, ISzAllocPtr alloc); + +typedef struct +{ + Byte *data; + size_t size; + size_t pos; +} CDynBuf; + +void DynBuf_Construct(CDynBuf *p); +void DynBuf_SeekToBeg(CDynBuf *p); +int DynBuf_Write(CDynBuf *p, const Byte *buf, size_t size, ISzAllocPtr alloc); +void DynBuf_Free(CDynBuf *p, ISzAllocPtr alloc); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/7zBuf2.c b/deps/LZMA-SDK/C/7zBuf2.c index 49b4343b6..208347416 100644 --- a/deps/LZMA-SDK/C/7zBuf2.c +++ b/deps/LZMA-SDK/C/7zBuf2.c @@ -1,52 +1,52 @@ -/* 7zBuf2.c -- Byte Buffer -2017-04-03 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include - -#include "7zBuf.h" - -void DynBuf_Construct(CDynBuf *p) -{ - p->data = 0; - p->size = 0; - p->pos = 0; -} - -void DynBuf_SeekToBeg(CDynBuf *p) -{ - p->pos = 0; -} - -int DynBuf_Write(CDynBuf *p, const Byte *buf, size_t size, ISzAllocPtr alloc) -{ - if (size > p->size - p->pos) - { - size_t newSize = p->pos + size; - Byte *data; - newSize += newSize / 4; - data = (Byte *)ISzAlloc_Alloc(alloc, newSize); - if (!data) - return 0; - p->size = newSize; - if (p->pos != 0) - memcpy(data, p->data, p->pos); - ISzAlloc_Free(alloc, p->data); - p->data = data; - } - if (size != 0) - { - memcpy(p->data + p->pos, buf, size); - p->pos += size; - } - return 1; -} - -void DynBuf_Free(CDynBuf *p, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, p->data); - p->data = 0; - p->size = 0; - p->pos = 0; -} +/* 7zBuf2.c -- Byte Buffer +2017-04-03 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#include "7zBuf.h" + +void DynBuf_Construct(CDynBuf *p) +{ + p->data = 0; + p->size = 0; + p->pos = 0; +} + +void DynBuf_SeekToBeg(CDynBuf *p) +{ + p->pos = 0; +} + +int DynBuf_Write(CDynBuf *p, const Byte *buf, size_t size, ISzAllocPtr alloc) +{ + if (size > p->size - p->pos) + { + size_t newSize = p->pos + size; + Byte *data; + newSize += newSize / 4; + data = (Byte *)ISzAlloc_Alloc(alloc, newSize); + if (!data) + return 0; + p->size = newSize; + if (p->pos != 0) + memcpy(data, p->data, p->pos); + ISzAlloc_Free(alloc, p->data); + p->data = data; + } + if (size != 0) + { + memcpy(p->data + p->pos, buf, size); + p->pos += size; + } + return 1; +} + +void DynBuf_Free(CDynBuf *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->data); + p->data = 0; + p->size = 0; + p->pos = 0; +} diff --git a/deps/LZMA-SDK/C/7zCrc.c b/deps/LZMA-SDK/C/7zCrc.c index c7ec353d6..f186324dd 100644 --- a/deps/LZMA-SDK/C/7zCrc.c +++ b/deps/LZMA-SDK/C/7zCrc.c @@ -1,322 +1,322 @@ -/* 7zCrc.c -- CRC32 init -2021-04-01 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "7zCrc.h" -#include "CpuArch.h" - -#define kCrcPoly 0xEDB88320 - -#ifdef MY_CPU_LE - #define CRC_NUM_TABLES 8 -#else - #define CRC_NUM_TABLES 9 - - #define CRC_UINT32_SWAP(v) ((v >> 24) | ((v >> 8) & 0xFF00) | ((v << 8) & 0xFF0000) | (v << 24)) - - UInt32 MY_FAST_CALL CrcUpdateT1_BeT4(UInt32 v, const void *data, size_t size, const UInt32 *table); - UInt32 MY_FAST_CALL CrcUpdateT1_BeT8(UInt32 v, const void *data, size_t size, const UInt32 *table); -#endif - -#ifndef MY_CPU_BE - UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const UInt32 *table); - UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const UInt32 *table); -#endif - -typedef UInt32 (MY_FAST_CALL *CRC_FUNC)(UInt32 v, const void *data, size_t size, const UInt32 *table); - -extern -CRC_FUNC g_CrcUpdateT4; -CRC_FUNC g_CrcUpdateT4; -extern -CRC_FUNC g_CrcUpdateT8; -CRC_FUNC g_CrcUpdateT8; -extern -CRC_FUNC g_CrcUpdateT0_32; -CRC_FUNC g_CrcUpdateT0_32; -extern -CRC_FUNC g_CrcUpdateT0_64; -CRC_FUNC g_CrcUpdateT0_64; -extern -CRC_FUNC g_CrcUpdate; -CRC_FUNC g_CrcUpdate; - -UInt32 g_CrcTable[256 * CRC_NUM_TABLES]; - -UInt32 MY_FAST_CALL CrcUpdate(UInt32 v, const void *data, size_t size) -{ - return g_CrcUpdate(v, data, size, g_CrcTable); -} - -UInt32 MY_FAST_CALL CrcCalc(const void *data, size_t size) -{ - return g_CrcUpdate(CRC_INIT_VAL, data, size, g_CrcTable) ^ CRC_INIT_VAL; -} - -#define CRC_UPDATE_BYTE_2(crc, b) (table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) - -UInt32 MY_FAST_CALL CrcUpdateT1(UInt32 v, const void *data, size_t size, const UInt32 *table); -UInt32 MY_FAST_CALL CrcUpdateT1(UInt32 v, const void *data, size_t size, const UInt32 *table) -{ - const Byte *p = (const Byte *)data; - const Byte *pEnd = p + size; - for (; p != pEnd; p++) - v = CRC_UPDATE_BYTE_2(v, *p); - return v; -} - - -/* ---------- hardware CRC ---------- */ - -#ifdef MY_CPU_LE - -#if defined(MY_CPU_ARM_OR_ARM64) - -// #pragma message("ARM*") - - #if defined(_MSC_VER) - #if defined(MY_CPU_ARM64) - #if (_MSC_VER >= 1910) - #define USE_ARM64_CRC - #endif - #endif - #elif (defined(__clang__) && (__clang_major__ >= 3)) \ - || (defined(__GNUC__) && (__GNUC__ > 4)) - #if !defined(__ARM_FEATURE_CRC32) - #define __ARM_FEATURE_CRC32 1 - #if (!defined(__clang__) || (__clang_major__ > 3)) // fix these numbers - #define ATTRIB_CRC __attribute__((__target__("arch=armv8-a+crc"))) - #endif - #endif - #if defined(__ARM_FEATURE_CRC32) - #define USE_ARM64_CRC - #include - #endif - #endif - -#else - -// no hardware CRC - -// #define USE_CRC_EMU - -#ifdef USE_CRC_EMU - -#pragma message("ARM64 CRC emulation") - -MY_FORCE_INLINE -UInt32 __crc32b(UInt32 v, UInt32 data) -{ - const UInt32 *table = g_CrcTable; - v = CRC_UPDATE_BYTE_2(v, (Byte)data); - return v; -} - -MY_FORCE_INLINE -UInt32 __crc32w(UInt32 v, UInt32 data) -{ - const UInt32 *table = g_CrcTable; - v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; - v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; - v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; - v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; - return v; -} - -MY_FORCE_INLINE -UInt32 __crc32d(UInt32 v, UInt64 data) -{ - const UInt32 *table = g_CrcTable; - v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; - v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; - v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; - v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; - v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; - v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; - v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; - v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; - return v; -} - -#endif // USE_CRC_EMU - -#endif // defined(MY_CPU_ARM64) && defined(MY_CPU_LE) - - - -#if defined(USE_ARM64_CRC) || defined(USE_CRC_EMU) - -#define T0_32_UNROLL_BYTES (4 * 4) -#define T0_64_UNROLL_BYTES (4 * 8) - -#ifndef ATTRIB_CRC -#define ATTRIB_CRC -#endif -// #pragma message("USE ARM HW CRC") - -ATTRIB_CRC -UInt32 MY_FAST_CALL CrcUpdateT0_32(UInt32 v, const void *data, size_t size, const UInt32 *table); -ATTRIB_CRC -UInt32 MY_FAST_CALL CrcUpdateT0_32(UInt32 v, const void *data, size_t size, const UInt32 *table) -{ - const Byte *p = (const Byte *)data; - UNUSED_VAR(table); - - for (; size != 0 && ((unsigned)(ptrdiff_t)p & (T0_32_UNROLL_BYTES - 1)) != 0; size--) - v = __crc32b(v, *p++); - - if (size >= T0_32_UNROLL_BYTES) - { - const Byte *lim = p + size; - size &= (T0_32_UNROLL_BYTES - 1); - lim -= size; - do - { - v = __crc32w(v, *(const UInt32 *)(const void *)(p)); - v = __crc32w(v, *(const UInt32 *)(const void *)(p + 4)); p += 2 * 4; - v = __crc32w(v, *(const UInt32 *)(const void *)(p)); - v = __crc32w(v, *(const UInt32 *)(const void *)(p + 4)); p += 2 * 4; - } - while (p != lim); - } - - for (; size != 0; size--) - v = __crc32b(v, *p++); - - return v; -} - -ATTRIB_CRC -UInt32 MY_FAST_CALL CrcUpdateT0_64(UInt32 v, const void *data, size_t size, const UInt32 *table); -ATTRIB_CRC -UInt32 MY_FAST_CALL CrcUpdateT0_64(UInt32 v, const void *data, size_t size, const UInt32 *table) -{ - const Byte *p = (const Byte *)data; - UNUSED_VAR(table); - - for (; size != 0 && ((unsigned)(ptrdiff_t)p & (T0_64_UNROLL_BYTES - 1)) != 0; size--) - v = __crc32b(v, *p++); - - if (size >= T0_64_UNROLL_BYTES) - { - const Byte *lim = p + size; - size &= (T0_64_UNROLL_BYTES - 1); - lim -= size; - do - { - v = __crc32d(v, *(const UInt64 *)(const void *)(p)); - v = __crc32d(v, *(const UInt64 *)(const void *)(p + 8)); p += 2 * 8; - v = __crc32d(v, *(const UInt64 *)(const void *)(p)); - v = __crc32d(v, *(const UInt64 *)(const void *)(p + 8)); p += 2 * 8; - } - while (p != lim); - } - - for (; size != 0; size--) - v = __crc32b(v, *p++); - - return v; -} - -#endif // defined(USE_ARM64_CRC) || defined(USE_CRC_EMU) - -#endif // MY_CPU_LE - - - - -void MY_FAST_CALL CrcGenerateTable() -{ - UInt32 i; - for (i = 0; i < 256; i++) - { - UInt32 r = i; - unsigned j; - for (j = 0; j < 8; j++) - r = (r >> 1) ^ (kCrcPoly & ((UInt32)0 - (r & 1))); - g_CrcTable[i] = r; - } - for (i = 256; i < 256 * CRC_NUM_TABLES; i++) - { - UInt32 r = g_CrcTable[(size_t)i - 256]; - g_CrcTable[i] = g_CrcTable[r & 0xFF] ^ (r >> 8); - } - - #if CRC_NUM_TABLES < 4 - - g_CrcUpdate = CrcUpdateT1; - - #else - - #ifdef MY_CPU_LE - - g_CrcUpdateT4 = CrcUpdateT4; - g_CrcUpdate = CrcUpdateT4; - - #if CRC_NUM_TABLES >= 8 - g_CrcUpdateT8 = CrcUpdateT8; - - #ifdef MY_CPU_X86_OR_AMD64 - if (!CPU_Is_InOrder()) - #endif - g_CrcUpdate = CrcUpdateT8; - #endif - - #else - { - #ifndef MY_CPU_BE - UInt32 k = 0x01020304; - const Byte *p = (const Byte *)&k; - if (p[0] == 4 && p[1] == 3) - { - g_CrcUpdateT4 = CrcUpdateT4; - g_CrcUpdate = CrcUpdateT4; - #if CRC_NUM_TABLES >= 8 - g_CrcUpdateT8 = CrcUpdateT8; - g_CrcUpdate = CrcUpdateT8; - #endif - } - else if (p[0] != 1 || p[1] != 2) - g_CrcUpdate = CrcUpdateT1; - else - #endif - { - for (i = 256 * CRC_NUM_TABLES - 1; i >= 256; i--) - { - UInt32 x = g_CrcTable[(size_t)i - 256]; - g_CrcTable[i] = CRC_UINT32_SWAP(x); - } - g_CrcUpdateT4 = CrcUpdateT1_BeT4; - g_CrcUpdate = CrcUpdateT1_BeT4; - #if CRC_NUM_TABLES >= 8 - g_CrcUpdateT8 = CrcUpdateT1_BeT8; - g_CrcUpdate = CrcUpdateT1_BeT8; - #endif - } - } - #endif - #endif - - #ifdef MY_CPU_LE - #ifdef USE_ARM64_CRC - if (CPU_IsSupported_CRC32()) - { - g_CrcUpdateT0_32 = CrcUpdateT0_32; - g_CrcUpdateT0_64 = CrcUpdateT0_64; - g_CrcUpdate = - #if defined(MY_CPU_ARM) - CrcUpdateT0_32; - #else - CrcUpdateT0_64; - #endif - } - #endif - - #ifdef USE_CRC_EMU - g_CrcUpdateT0_32 = CrcUpdateT0_32; - g_CrcUpdateT0_64 = CrcUpdateT0_64; - g_CrcUpdate = CrcUpdateT0_64; - #endif - #endif -} +/* 7zCrc.c -- CRC32 init +2021-04-01 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "7zCrc.h" +#include "CpuArch.h" + +#define kCrcPoly 0xEDB88320 + +#ifdef MY_CPU_LE + #define CRC_NUM_TABLES 8 +#else + #define CRC_NUM_TABLES 9 + + #define CRC_UINT32_SWAP(v) ((v >> 24) | ((v >> 8) & 0xFF00) | ((v << 8) & 0xFF0000) | (v << 24)) + + UInt32 MY_FAST_CALL CrcUpdateT1_BeT4(UInt32 v, const void *data, size_t size, const UInt32 *table); + UInt32 MY_FAST_CALL CrcUpdateT1_BeT8(UInt32 v, const void *data, size_t size, const UInt32 *table); +#endif + +#ifndef MY_CPU_BE + UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const UInt32 *table); + UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const UInt32 *table); +#endif + +typedef UInt32 (MY_FAST_CALL *CRC_FUNC)(UInt32 v, const void *data, size_t size, const UInt32 *table); + +extern +CRC_FUNC g_CrcUpdateT4; +CRC_FUNC g_CrcUpdateT4; +extern +CRC_FUNC g_CrcUpdateT8; +CRC_FUNC g_CrcUpdateT8; +extern +CRC_FUNC g_CrcUpdateT0_32; +CRC_FUNC g_CrcUpdateT0_32; +extern +CRC_FUNC g_CrcUpdateT0_64; +CRC_FUNC g_CrcUpdateT0_64; +extern +CRC_FUNC g_CrcUpdate; +CRC_FUNC g_CrcUpdate; + +UInt32 g_CrcTable[256 * CRC_NUM_TABLES]; + +UInt32 MY_FAST_CALL CrcUpdate(UInt32 v, const void *data, size_t size) +{ + return g_CrcUpdate(v, data, size, g_CrcTable); +} + +UInt32 MY_FAST_CALL CrcCalc(const void *data, size_t size) +{ + return g_CrcUpdate(CRC_INIT_VAL, data, size, g_CrcTable) ^ CRC_INIT_VAL; +} + +#define CRC_UPDATE_BYTE_2(crc, b) (table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) + +UInt32 MY_FAST_CALL CrcUpdateT1(UInt32 v, const void *data, size_t size, const UInt32 *table); +UInt32 MY_FAST_CALL CrcUpdateT1(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + const Byte *pEnd = p + size; + for (; p != pEnd; p++) + v = CRC_UPDATE_BYTE_2(v, *p); + return v; +} + + +/* ---------- hardware CRC ---------- */ + +#ifdef MY_CPU_LE + +#if defined(MY_CPU_ARM_OR_ARM64) + +// #pragma message("ARM*") + + #if defined(_MSC_VER) + #if defined(MY_CPU_ARM64) + #if (_MSC_VER >= 1910) + #define USE_ARM64_CRC + #endif + #endif + #elif (defined(__clang__) && (__clang_major__ >= 3)) \ + || (defined(__GNUC__) && (__GNUC__ > 4)) + #if !defined(__ARM_FEATURE_CRC32) + #define __ARM_FEATURE_CRC32 1 + #if (!defined(__clang__) || (__clang_major__ > 3)) // fix these numbers + #define ATTRIB_CRC __attribute__((__target__("arch=armv8-a+crc"))) + #endif + #endif + #if defined(__ARM_FEATURE_CRC32) + #define USE_ARM64_CRC + #include + #endif + #endif + +#else + +// no hardware CRC + +// #define USE_CRC_EMU + +#ifdef USE_CRC_EMU + +#pragma message("ARM64 CRC emulation") + +MY_FORCE_INLINE +UInt32 __crc32b(UInt32 v, UInt32 data) +{ + const UInt32 *table = g_CrcTable; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); + return v; +} + +MY_FORCE_INLINE +UInt32 __crc32w(UInt32 v, UInt32 data) +{ + const UInt32 *table = g_CrcTable; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + return v; +} + +MY_FORCE_INLINE +UInt32 __crc32d(UInt32 v, UInt64 data) +{ + const UInt32 *table = g_CrcTable; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + return v; +} + +#endif // USE_CRC_EMU + +#endif // defined(MY_CPU_ARM64) && defined(MY_CPU_LE) + + + +#if defined(USE_ARM64_CRC) || defined(USE_CRC_EMU) + +#define T0_32_UNROLL_BYTES (4 * 4) +#define T0_64_UNROLL_BYTES (4 * 8) + +#ifndef ATTRIB_CRC +#define ATTRIB_CRC +#endif +// #pragma message("USE ARM HW CRC") + +ATTRIB_CRC +UInt32 MY_FAST_CALL CrcUpdateT0_32(UInt32 v, const void *data, size_t size, const UInt32 *table); +ATTRIB_CRC +UInt32 MY_FAST_CALL CrcUpdateT0_32(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + UNUSED_VAR(table); + + for (; size != 0 && ((unsigned)(ptrdiff_t)p & (T0_32_UNROLL_BYTES - 1)) != 0; size--) + v = __crc32b(v, *p++); + + if (size >= T0_32_UNROLL_BYTES) + { + const Byte *lim = p + size; + size &= (T0_32_UNROLL_BYTES - 1); + lim -= size; + do + { + v = __crc32w(v, *(const UInt32 *)(const void *)(p)); + v = __crc32w(v, *(const UInt32 *)(const void *)(p + 4)); p += 2 * 4; + v = __crc32w(v, *(const UInt32 *)(const void *)(p)); + v = __crc32w(v, *(const UInt32 *)(const void *)(p + 4)); p += 2 * 4; + } + while (p != lim); + } + + for (; size != 0; size--) + v = __crc32b(v, *p++); + + return v; +} + +ATTRIB_CRC +UInt32 MY_FAST_CALL CrcUpdateT0_64(UInt32 v, const void *data, size_t size, const UInt32 *table); +ATTRIB_CRC +UInt32 MY_FAST_CALL CrcUpdateT0_64(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + UNUSED_VAR(table); + + for (; size != 0 && ((unsigned)(ptrdiff_t)p & (T0_64_UNROLL_BYTES - 1)) != 0; size--) + v = __crc32b(v, *p++); + + if (size >= T0_64_UNROLL_BYTES) + { + const Byte *lim = p + size; + size &= (T0_64_UNROLL_BYTES - 1); + lim -= size; + do + { + v = __crc32d(v, *(const UInt64 *)(const void *)(p)); + v = __crc32d(v, *(const UInt64 *)(const void *)(p + 8)); p += 2 * 8; + v = __crc32d(v, *(const UInt64 *)(const void *)(p)); + v = __crc32d(v, *(const UInt64 *)(const void *)(p + 8)); p += 2 * 8; + } + while (p != lim); + } + + for (; size != 0; size--) + v = __crc32b(v, *p++); + + return v; +} + +#endif // defined(USE_ARM64_CRC) || defined(USE_CRC_EMU) + +#endif // MY_CPU_LE + + + + +void MY_FAST_CALL CrcGenerateTable() +{ + UInt32 i; + for (i = 0; i < 256; i++) + { + UInt32 r = i; + unsigned j; + for (j = 0; j < 8; j++) + r = (r >> 1) ^ (kCrcPoly & ((UInt32)0 - (r & 1))); + g_CrcTable[i] = r; + } + for (i = 256; i < 256 * CRC_NUM_TABLES; i++) + { + UInt32 r = g_CrcTable[(size_t)i - 256]; + g_CrcTable[i] = g_CrcTable[r & 0xFF] ^ (r >> 8); + } + + #if CRC_NUM_TABLES < 4 + + g_CrcUpdate = CrcUpdateT1; + + #else + + #ifdef MY_CPU_LE + + g_CrcUpdateT4 = CrcUpdateT4; + g_CrcUpdate = CrcUpdateT4; + + #if CRC_NUM_TABLES >= 8 + g_CrcUpdateT8 = CrcUpdateT8; + + #ifdef MY_CPU_X86_OR_AMD64 + if (!CPU_Is_InOrder()) + #endif + g_CrcUpdate = CrcUpdateT8; + #endif + + #else + { + #ifndef MY_CPU_BE + UInt32 k = 0x01020304; + const Byte *p = (const Byte *)&k; + if (p[0] == 4 && p[1] == 3) + { + g_CrcUpdateT4 = CrcUpdateT4; + g_CrcUpdate = CrcUpdateT4; + #if CRC_NUM_TABLES >= 8 + g_CrcUpdateT8 = CrcUpdateT8; + g_CrcUpdate = CrcUpdateT8; + #endif + } + else if (p[0] != 1 || p[1] != 2) + g_CrcUpdate = CrcUpdateT1; + else + #endif + { + for (i = 256 * CRC_NUM_TABLES - 1; i >= 256; i--) + { + UInt32 x = g_CrcTable[(size_t)i - 256]; + g_CrcTable[i] = CRC_UINT32_SWAP(x); + } + g_CrcUpdateT4 = CrcUpdateT1_BeT4; + g_CrcUpdate = CrcUpdateT1_BeT4; + #if CRC_NUM_TABLES >= 8 + g_CrcUpdateT8 = CrcUpdateT1_BeT8; + g_CrcUpdate = CrcUpdateT1_BeT8; + #endif + } + } + #endif + #endif + + #ifdef MY_CPU_LE + #ifdef USE_ARM64_CRC + if (CPU_IsSupported_CRC32()) + { + g_CrcUpdateT0_32 = CrcUpdateT0_32; + g_CrcUpdateT0_64 = CrcUpdateT0_64; + g_CrcUpdate = + #if defined(MY_CPU_ARM) + CrcUpdateT0_32; + #else + CrcUpdateT0_64; + #endif + } + #endif + + #ifdef USE_CRC_EMU + g_CrcUpdateT0_32 = CrcUpdateT0_32; + g_CrcUpdateT0_64 = CrcUpdateT0_64; + g_CrcUpdate = CrcUpdateT0_64; + #endif + #endif +} diff --git a/deps/LZMA-SDK/C/7zCrc.h b/deps/LZMA-SDK/C/7zCrc.h index 3b0459402..8fd579587 100644 --- a/deps/LZMA-SDK/C/7zCrc.h +++ b/deps/LZMA-SDK/C/7zCrc.h @@ -1,25 +1,25 @@ -/* 7zCrc.h -- CRC32 calculation -2013-01-18 : Igor Pavlov : Public domain */ - -#ifndef __7Z_CRC_H -#define __7Z_CRC_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -extern UInt32 g_CrcTable[]; - -/* Call CrcGenerateTable one time before other CRC functions */ -void MY_FAST_CALL CrcGenerateTable(void); - -#define CRC_INIT_VAL 0xFFFFFFFF -#define CRC_GET_DIGEST(crc) ((crc) ^ CRC_INIT_VAL) -#define CRC_UPDATE_BYTE(crc, b) (g_CrcTable[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) - -UInt32 MY_FAST_CALL CrcUpdate(UInt32 crc, const void *data, size_t size); -UInt32 MY_FAST_CALL CrcCalc(const void *data, size_t size); - -EXTERN_C_END - -#endif +/* 7zCrc.h -- CRC32 calculation +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __7Z_CRC_H +#define __7Z_CRC_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +extern UInt32 g_CrcTable[]; + +/* Call CrcGenerateTable one time before other CRC functions */ +void MY_FAST_CALL CrcGenerateTable(void); + +#define CRC_INIT_VAL 0xFFFFFFFF +#define CRC_GET_DIGEST(crc) ((crc) ^ CRC_INIT_VAL) +#define CRC_UPDATE_BYTE(crc, b) (g_CrcTable[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) + +UInt32 MY_FAST_CALL CrcUpdate(UInt32 crc, const void *data, size_t size); +UInt32 MY_FAST_CALL CrcCalc(const void *data, size_t size); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/7zCrcOpt.c b/deps/LZMA-SDK/C/7zCrcOpt.c index efaa7ab9d..69fad9ca2 100644 --- a/deps/LZMA-SDK/C/7zCrcOpt.c +++ b/deps/LZMA-SDK/C/7zCrcOpt.c @@ -1,117 +1,117 @@ -/* 7zCrcOpt.c -- CRC32 calculation -2021-02-09 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "CpuArch.h" - -#ifndef MY_CPU_BE - -#define CRC_UPDATE_BYTE_2(crc, b) (table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) - -UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const UInt32 *table); -UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const UInt32 *table) -{ - const Byte *p = (const Byte *)data; - for (; size > 0 && ((unsigned)(ptrdiff_t)p & 3) != 0; size--, p++) - v = CRC_UPDATE_BYTE_2(v, *p); - for (; size >= 4; size -= 4, p += 4) - { - v ^= *(const UInt32 *)(const void *)p; - v = - (table + 0x300)[((v ) & 0xFF)] - ^ (table + 0x200)[((v >> 8) & 0xFF)] - ^ (table + 0x100)[((v >> 16) & 0xFF)] - ^ (table + 0x000)[((v >> 24))]; - } - for (; size > 0; size--, p++) - v = CRC_UPDATE_BYTE_2(v, *p); - return v; -} - -UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const UInt32 *table); -UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const UInt32 *table) -{ - const Byte *p = (const Byte *)data; - for (; size > 0 && ((unsigned)(ptrdiff_t)p & 7) != 0; size--, p++) - v = CRC_UPDATE_BYTE_2(v, *p); - for (; size >= 8; size -= 8, p += 8) - { - UInt32 d; - v ^= *(const UInt32 *)(const void *)p; - v = - (table + 0x700)[((v ) & 0xFF)] - ^ (table + 0x600)[((v >> 8) & 0xFF)] - ^ (table + 0x500)[((v >> 16) & 0xFF)] - ^ (table + 0x400)[((v >> 24))]; - d = *((const UInt32 *)(const void *)p + 1); - v ^= - (table + 0x300)[((d ) & 0xFF)] - ^ (table + 0x200)[((d >> 8) & 0xFF)] - ^ (table + 0x100)[((d >> 16) & 0xFF)] - ^ (table + 0x000)[((d >> 24))]; - } - for (; size > 0; size--, p++) - v = CRC_UPDATE_BYTE_2(v, *p); - return v; -} - -#endif - - -#ifndef MY_CPU_LE - -#define CRC_UINT32_SWAP(v) ((v >> 24) | ((v >> 8) & 0xFF00) | ((v << 8) & 0xFF0000) | (v << 24)) - -#define CRC_UPDATE_BYTE_2_BE(crc, b) (table[(((crc) >> 24) ^ (b))] ^ ((crc) << 8)) - -UInt32 MY_FAST_CALL CrcUpdateT1_BeT4(UInt32 v, const void *data, size_t size, const UInt32 *table) -{ - const Byte *p = (const Byte *)data; - table += 0x100; - v = CRC_UINT32_SWAP(v); - for (; size > 0 && ((unsigned)(ptrdiff_t)p & 3) != 0; size--, p++) - v = CRC_UPDATE_BYTE_2_BE(v, *p); - for (; size >= 4; size -= 4, p += 4) - { - v ^= *(const UInt32 *)(const void *)p; - v = - (table + 0x000)[((v ) & 0xFF)] - ^ (table + 0x100)[((v >> 8) & 0xFF)] - ^ (table + 0x200)[((v >> 16) & 0xFF)] - ^ (table + 0x300)[((v >> 24))]; - } - for (; size > 0; size--, p++) - v = CRC_UPDATE_BYTE_2_BE(v, *p); - return CRC_UINT32_SWAP(v); -} - -UInt32 MY_FAST_CALL CrcUpdateT1_BeT8(UInt32 v, const void *data, size_t size, const UInt32 *table) -{ - const Byte *p = (const Byte *)data; - table += 0x100; - v = CRC_UINT32_SWAP(v); - for (; size > 0 && ((unsigned)(ptrdiff_t)p & 7) != 0; size--, p++) - v = CRC_UPDATE_BYTE_2_BE(v, *p); - for (; size >= 8; size -= 8, p += 8) - { - UInt32 d; - v ^= *(const UInt32 *)(const void *)p; - v = - (table + 0x400)[((v ) & 0xFF)] - ^ (table + 0x500)[((v >> 8) & 0xFF)] - ^ (table + 0x600)[((v >> 16) & 0xFF)] - ^ (table + 0x700)[((v >> 24))]; - d = *((const UInt32 *)(const void *)p + 1); - v ^= - (table + 0x000)[((d ) & 0xFF)] - ^ (table + 0x100)[((d >> 8) & 0xFF)] - ^ (table + 0x200)[((d >> 16) & 0xFF)] - ^ (table + 0x300)[((d >> 24))]; - } - for (; size > 0; size--, p++) - v = CRC_UPDATE_BYTE_2_BE(v, *p); - return CRC_UINT32_SWAP(v); -} - -#endif +/* 7zCrcOpt.c -- CRC32 calculation +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" + +#ifndef MY_CPU_BE + +#define CRC_UPDATE_BYTE_2(crc, b) (table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) + +UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const UInt32 *table); +UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + for (; size > 0 && ((unsigned)(ptrdiff_t)p & 3) != 0; size--, p++) + v = CRC_UPDATE_BYTE_2(v, *p); + for (; size >= 4; size -= 4, p += 4) + { + v ^= *(const UInt32 *)(const void *)p; + v = + (table + 0x300)[((v ) & 0xFF)] + ^ (table + 0x200)[((v >> 8) & 0xFF)] + ^ (table + 0x100)[((v >> 16) & 0xFF)] + ^ (table + 0x000)[((v >> 24))]; + } + for (; size > 0; size--, p++) + v = CRC_UPDATE_BYTE_2(v, *p); + return v; +} + +UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const UInt32 *table); +UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + for (; size > 0 && ((unsigned)(ptrdiff_t)p & 7) != 0; size--, p++) + v = CRC_UPDATE_BYTE_2(v, *p); + for (; size >= 8; size -= 8, p += 8) + { + UInt32 d; + v ^= *(const UInt32 *)(const void *)p; + v = + (table + 0x700)[((v ) & 0xFF)] + ^ (table + 0x600)[((v >> 8) & 0xFF)] + ^ (table + 0x500)[((v >> 16) & 0xFF)] + ^ (table + 0x400)[((v >> 24))]; + d = *((const UInt32 *)(const void *)p + 1); + v ^= + (table + 0x300)[((d ) & 0xFF)] + ^ (table + 0x200)[((d >> 8) & 0xFF)] + ^ (table + 0x100)[((d >> 16) & 0xFF)] + ^ (table + 0x000)[((d >> 24))]; + } + for (; size > 0; size--, p++) + v = CRC_UPDATE_BYTE_2(v, *p); + return v; +} + +#endif + + +#ifndef MY_CPU_LE + +#define CRC_UINT32_SWAP(v) ((v >> 24) | ((v >> 8) & 0xFF00) | ((v << 8) & 0xFF0000) | (v << 24)) + +#define CRC_UPDATE_BYTE_2_BE(crc, b) (table[(((crc) >> 24) ^ (b))] ^ ((crc) << 8)) + +UInt32 MY_FAST_CALL CrcUpdateT1_BeT4(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + table += 0x100; + v = CRC_UINT32_SWAP(v); + for (; size > 0 && ((unsigned)(ptrdiff_t)p & 3) != 0; size--, p++) + v = CRC_UPDATE_BYTE_2_BE(v, *p); + for (; size >= 4; size -= 4, p += 4) + { + v ^= *(const UInt32 *)(const void *)p; + v = + (table + 0x000)[((v ) & 0xFF)] + ^ (table + 0x100)[((v >> 8) & 0xFF)] + ^ (table + 0x200)[((v >> 16) & 0xFF)] + ^ (table + 0x300)[((v >> 24))]; + } + for (; size > 0; size--, p++) + v = CRC_UPDATE_BYTE_2_BE(v, *p); + return CRC_UINT32_SWAP(v); +} + +UInt32 MY_FAST_CALL CrcUpdateT1_BeT8(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + table += 0x100; + v = CRC_UINT32_SWAP(v); + for (; size > 0 && ((unsigned)(ptrdiff_t)p & 7) != 0; size--, p++) + v = CRC_UPDATE_BYTE_2_BE(v, *p); + for (; size >= 8; size -= 8, p += 8) + { + UInt32 d; + v ^= *(const UInt32 *)(const void *)p; + v = + (table + 0x400)[((v ) & 0xFF)] + ^ (table + 0x500)[((v >> 8) & 0xFF)] + ^ (table + 0x600)[((v >> 16) & 0xFF)] + ^ (table + 0x700)[((v >> 24))]; + d = *((const UInt32 *)(const void *)p + 1); + v ^= + (table + 0x000)[((d ) & 0xFF)] + ^ (table + 0x100)[((d >> 8) & 0xFF)] + ^ (table + 0x200)[((d >> 16) & 0xFF)] + ^ (table + 0x300)[((d >> 24))]; + } + for (; size > 0; size--, p++) + v = CRC_UPDATE_BYTE_2_BE(v, *p); + return CRC_UINT32_SWAP(v); +} + +#endif diff --git a/deps/LZMA-SDK/C/7zDec.c b/deps/LZMA-SDK/C/7zDec.c index 83e37d166..fbfd016e1 100644 --- a/deps/LZMA-SDK/C/7zDec.c +++ b/deps/LZMA-SDK/C/7zDec.c @@ -1,600 +1,600 @@ -/* 7zDec.c -- Decoding from 7z folder -2021-02-09 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include - -/* #define _7ZIP_PPMD_SUPPPORT */ - -#include "7z.h" -#include "7zCrc.h" - -#include "Bcj2.h" -#include "Bra.h" -#include "CpuArch.h" -#include "Delta.h" -#include "LzmaDec.h" -#include "Lzma2Dec.h" -#ifdef _7ZIP_PPMD_SUPPPORT -#include "Ppmd7.h" -#endif - -#define k_Copy 0 -#ifndef _7Z_NO_METHOD_LZMA2 -#define k_LZMA2 0x21 -#endif -#define k_LZMA 0x30101 -#define k_BCJ2 0x303011B -#ifndef _7Z_NO_METHODS_FILTERS -#define k_Delta 3 -#define k_BCJ 0x3030103 -#define k_PPC 0x3030205 -#define k_IA64 0x3030401 -#define k_ARM 0x3030501 -#define k_ARMT 0x3030701 -#define k_SPARC 0x3030805 -#endif - -#ifdef _7ZIP_PPMD_SUPPPORT - -#define k_PPMD 0x30401 - -typedef struct -{ - IByteIn vt; - const Byte *cur; - const Byte *end; - const Byte *begin; - UInt64 processed; - BoolInt extra; - SRes res; - const ILookInStream *inStream; -} CByteInToLook; - -static Byte ReadByte(const IByteIn *pp) -{ - CByteInToLook *p = CONTAINER_FROM_VTBL(pp, CByteInToLook, vt); - if (p->cur != p->end) - return *p->cur++; - if (p->res == SZ_OK) - { - size_t size = (size_t)(p->cur - p->begin); - p->processed += size; - p->res = ILookInStream_Skip(p->inStream, size); - size = (1 << 25); - p->res = ILookInStream_Look(p->inStream, (const void **)&p->begin, &size); - p->cur = p->begin; - p->end = p->begin + size; - if (size != 0) - return *p->cur++;; - } - p->extra = True; - return 0; -} - -static SRes SzDecodePpmd(const Byte *props, unsigned propsSize, UInt64 inSize, const ILookInStream *inStream, - Byte *outBuffer, SizeT outSize, ISzAllocPtr allocMain) -{ - CPpmd7 ppmd; - CByteInToLook s; - SRes res = SZ_OK; - - s.vt.Read = ReadByte; - s.inStream = inStream; - s.begin = s.end = s.cur = NULL; - s.extra = False; - s.res = SZ_OK; - s.processed = 0; - - if (propsSize != 5) - return SZ_ERROR_UNSUPPORTED; - - { - unsigned order = props[0]; - UInt32 memSize = GetUi32(props + 1); - if (order < PPMD7_MIN_ORDER || - order > PPMD7_MAX_ORDER || - memSize < PPMD7_MIN_MEM_SIZE || - memSize > PPMD7_MAX_MEM_SIZE) - return SZ_ERROR_UNSUPPORTED; - Ppmd7_Construct(&ppmd); - if (!Ppmd7_Alloc(&ppmd, memSize, allocMain)) - return SZ_ERROR_MEM; - Ppmd7_Init(&ppmd, order); - } - { - ppmd.rc.dec.Stream = &s.vt; - if (!Ppmd7z_RangeDec_Init(&ppmd.rc.dec)) - res = SZ_ERROR_DATA; - else if (!s.extra) - { - Byte *buf = outBuffer; - const Byte *lim = buf + outSize; - for (; buf != lim; buf++) - { - int sym = Ppmd7z_DecodeSymbol(&ppmd); - if (s.extra || sym < 0) - break; - *buf = (Byte)sym; - } - if (buf != lim) - res = SZ_ERROR_DATA; - else if (!Ppmd7z_RangeDec_IsFinishedOK(&ppmd.rc.dec)) - { - /* if (Ppmd7z_DecodeSymbol(&ppmd) != PPMD7_SYM_END || !Ppmd7z_RangeDec_IsFinishedOK(&ppmd.rc.dec)) */ - res = SZ_ERROR_DATA; - } - } - if (s.extra) - res = (s.res != SZ_OK ? s.res : SZ_ERROR_DATA); - else if (s.processed + (size_t)(s.cur - s.begin) != inSize) - res = SZ_ERROR_DATA; - } - Ppmd7_Free(&ppmd, allocMain); - return res; -} - -#endif - - -static SRes SzDecodeLzma(const Byte *props, unsigned propsSize, UInt64 inSize, ILookInStream *inStream, - Byte *outBuffer, SizeT outSize, ISzAllocPtr allocMain) -{ - CLzmaDec state; - SRes res = SZ_OK; - - LzmaDec_Construct(&state); - RINOK(LzmaDec_AllocateProbs(&state, props, propsSize, allocMain)); - state.dic = outBuffer; - state.dicBufSize = outSize; - LzmaDec_Init(&state); - - for (;;) - { - const void *inBuf = NULL; - size_t lookahead = (1 << 18); - if (lookahead > inSize) - lookahead = (size_t)inSize; - res = ILookInStream_Look(inStream, &inBuf, &lookahead); - if (res != SZ_OK) - break; - - { - SizeT inProcessed = (SizeT)lookahead, dicPos = state.dicPos; - ELzmaStatus status; - res = LzmaDec_DecodeToDic(&state, outSize, (const Byte *)inBuf, &inProcessed, LZMA_FINISH_END, &status); - lookahead -= inProcessed; - inSize -= inProcessed; - if (res != SZ_OK) - break; - - if (status == LZMA_STATUS_FINISHED_WITH_MARK) - { - if (outSize != state.dicPos || inSize != 0) - res = SZ_ERROR_DATA; - break; - } - - if (outSize == state.dicPos && inSize == 0 && status == LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK) - break; - - if (inProcessed == 0 && dicPos == state.dicPos) - { - res = SZ_ERROR_DATA; - break; - } - - res = ILookInStream_Skip(inStream, inProcessed); - if (res != SZ_OK) - break; - } - } - - LzmaDec_FreeProbs(&state, allocMain); - return res; -} - - -#ifndef _7Z_NO_METHOD_LZMA2 - -static SRes SzDecodeLzma2(const Byte *props, unsigned propsSize, UInt64 inSize, ILookInStream *inStream, - Byte *outBuffer, SizeT outSize, ISzAllocPtr allocMain) -{ - CLzma2Dec state; - SRes res = SZ_OK; - - Lzma2Dec_Construct(&state); - if (propsSize != 1) - return SZ_ERROR_DATA; - RINOK(Lzma2Dec_AllocateProbs(&state, props[0], allocMain)); - state.decoder.dic = outBuffer; - state.decoder.dicBufSize = outSize; - Lzma2Dec_Init(&state); - - for (;;) - { - const void *inBuf = NULL; - size_t lookahead = (1 << 18); - if (lookahead > inSize) - lookahead = (size_t)inSize; - res = ILookInStream_Look(inStream, &inBuf, &lookahead); - if (res != SZ_OK) - break; - - { - SizeT inProcessed = (SizeT)lookahead, dicPos = state.decoder.dicPos; - ELzmaStatus status; - res = Lzma2Dec_DecodeToDic(&state, outSize, (const Byte *)inBuf, &inProcessed, LZMA_FINISH_END, &status); - lookahead -= inProcessed; - inSize -= inProcessed; - if (res != SZ_OK) - break; - - if (status == LZMA_STATUS_FINISHED_WITH_MARK) - { - if (outSize != state.decoder.dicPos || inSize != 0) - res = SZ_ERROR_DATA; - break; - } - - if (inProcessed == 0 && dicPos == state.decoder.dicPos) - { - res = SZ_ERROR_DATA; - break; - } - - res = ILookInStream_Skip(inStream, inProcessed); - if (res != SZ_OK) - break; - } - } - - Lzma2Dec_FreeProbs(&state, allocMain); - return res; -} - -#endif - - -static SRes SzDecodeCopy(UInt64 inSize, ILookInStream *inStream, Byte *outBuffer) -{ - while (inSize > 0) - { - const void *inBuf; - size_t curSize = (1 << 18); - if (curSize > inSize) - curSize = (size_t)inSize; - RINOK(ILookInStream_Look(inStream, &inBuf, &curSize)); - if (curSize == 0) - return SZ_ERROR_INPUT_EOF; - memcpy(outBuffer, inBuf, curSize); - outBuffer += curSize; - inSize -= curSize; - RINOK(ILookInStream_Skip(inStream, curSize)); - } - return SZ_OK; -} - -static BoolInt IS_MAIN_METHOD(UInt32 m) -{ - switch (m) - { - case k_Copy: - case k_LZMA: - #ifndef _7Z_NO_METHOD_LZMA2 - case k_LZMA2: - #endif - #ifdef _7ZIP_PPMD_SUPPPORT - case k_PPMD: - #endif - return True; - } - return False; -} - -static BoolInt IS_SUPPORTED_CODER(const CSzCoderInfo *c) -{ - return - c->NumStreams == 1 - /* && c->MethodID <= (UInt32)0xFFFFFFFF */ - && IS_MAIN_METHOD((UInt32)c->MethodID); -} - -#define IS_BCJ2(c) ((c)->MethodID == k_BCJ2 && (c)->NumStreams == 4) - -static SRes CheckSupportedFolder(const CSzFolder *f) -{ - if (f->NumCoders < 1 || f->NumCoders > 4) - return SZ_ERROR_UNSUPPORTED; - if (!IS_SUPPORTED_CODER(&f->Coders[0])) - return SZ_ERROR_UNSUPPORTED; - if (f->NumCoders == 1) - { - if (f->NumPackStreams != 1 || f->PackStreams[0] != 0 || f->NumBonds != 0) - return SZ_ERROR_UNSUPPORTED; - return SZ_OK; - } - - - #ifndef _7Z_NO_METHODS_FILTERS - - if (f->NumCoders == 2) - { - const CSzCoderInfo *c = &f->Coders[1]; - if ( - /* c->MethodID > (UInt32)0xFFFFFFFF || */ - c->NumStreams != 1 - || f->NumPackStreams != 1 - || f->PackStreams[0] != 0 - || f->NumBonds != 1 - || f->Bonds[0].InIndex != 1 - || f->Bonds[0].OutIndex != 0) - return SZ_ERROR_UNSUPPORTED; - switch ((UInt32)c->MethodID) - { - case k_Delta: - case k_BCJ: - case k_PPC: - case k_IA64: - case k_SPARC: - case k_ARM: - case k_ARMT: - break; - default: - return SZ_ERROR_UNSUPPORTED; - } - return SZ_OK; - } - - #endif - - - if (f->NumCoders == 4) - { - if (!IS_SUPPORTED_CODER(&f->Coders[1]) - || !IS_SUPPORTED_CODER(&f->Coders[2]) - || !IS_BCJ2(&f->Coders[3])) - return SZ_ERROR_UNSUPPORTED; - if (f->NumPackStreams != 4 - || f->PackStreams[0] != 2 - || f->PackStreams[1] != 6 - || f->PackStreams[2] != 1 - || f->PackStreams[3] != 0 - || f->NumBonds != 3 - || f->Bonds[0].InIndex != 5 || f->Bonds[0].OutIndex != 0 - || f->Bonds[1].InIndex != 4 || f->Bonds[1].OutIndex != 1 - || f->Bonds[2].InIndex != 3 || f->Bonds[2].OutIndex != 2) - return SZ_ERROR_UNSUPPORTED; - return SZ_OK; - } - - return SZ_ERROR_UNSUPPORTED; -} - -#ifndef _7Z_NO_METHODS_FILTERS -#define CASE_BRA_CONV(isa) case k_ ## isa: isa ## _Convert(outBuffer, outSize, 0, 0); break; -#endif - -static SRes SzFolder_Decode2(const CSzFolder *folder, - const Byte *propsData, - const UInt64 *unpackSizes, - const UInt64 *packPositions, - ILookInStream *inStream, UInt64 startPos, - Byte *outBuffer, SizeT outSize, ISzAllocPtr allocMain, - Byte *tempBuf[]) -{ - UInt32 ci; - SizeT tempSizes[3] = { 0, 0, 0}; - SizeT tempSize3 = 0; - Byte *tempBuf3 = 0; - - RINOK(CheckSupportedFolder(folder)); - - for (ci = 0; ci < folder->NumCoders; ci++) - { - const CSzCoderInfo *coder = &folder->Coders[ci]; - - if (IS_MAIN_METHOD((UInt32)coder->MethodID)) - { - UInt32 si = 0; - UInt64 offset; - UInt64 inSize; - Byte *outBufCur = outBuffer; - SizeT outSizeCur = outSize; - if (folder->NumCoders == 4) - { - UInt32 indices[] = { 3, 2, 0 }; - UInt64 unpackSize = unpackSizes[ci]; - si = indices[ci]; - if (ci < 2) - { - Byte *temp; - outSizeCur = (SizeT)unpackSize; - if (outSizeCur != unpackSize) - return SZ_ERROR_MEM; - temp = (Byte *)ISzAlloc_Alloc(allocMain, outSizeCur); - if (!temp && outSizeCur != 0) - return SZ_ERROR_MEM; - outBufCur = tempBuf[1 - ci] = temp; - tempSizes[1 - ci] = outSizeCur; - } - else if (ci == 2) - { - if (unpackSize > outSize) /* check it */ - return SZ_ERROR_PARAM; - tempBuf3 = outBufCur = outBuffer + (outSize - (size_t)unpackSize); - tempSize3 = outSizeCur = (SizeT)unpackSize; - } - else - return SZ_ERROR_UNSUPPORTED; - } - offset = packPositions[si]; - inSize = packPositions[(size_t)si + 1] - offset; - RINOK(LookInStream_SeekTo(inStream, startPos + offset)); - - if (coder->MethodID == k_Copy) - { - if (inSize != outSizeCur) /* check it */ - return SZ_ERROR_DATA; - RINOK(SzDecodeCopy(inSize, inStream, outBufCur)); - } - else if (coder->MethodID == k_LZMA) - { - RINOK(SzDecodeLzma(propsData + coder->PropsOffset, coder->PropsSize, inSize, inStream, outBufCur, outSizeCur, allocMain)); - } - #ifndef _7Z_NO_METHOD_LZMA2 - else if (coder->MethodID == k_LZMA2) - { - RINOK(SzDecodeLzma2(propsData + coder->PropsOffset, coder->PropsSize, inSize, inStream, outBufCur, outSizeCur, allocMain)); - } - #endif - #ifdef _7ZIP_PPMD_SUPPPORT - else if (coder->MethodID == k_PPMD) - { - RINOK(SzDecodePpmd(propsData + coder->PropsOffset, coder->PropsSize, inSize, inStream, outBufCur, outSizeCur, allocMain)); - } - #endif - else - return SZ_ERROR_UNSUPPORTED; - } - else if (coder->MethodID == k_BCJ2) - { - UInt64 offset = packPositions[1]; - UInt64 s3Size = packPositions[2] - offset; - - if (ci != 3) - return SZ_ERROR_UNSUPPORTED; - - tempSizes[2] = (SizeT)s3Size; - if (tempSizes[2] != s3Size) - return SZ_ERROR_MEM; - tempBuf[2] = (Byte *)ISzAlloc_Alloc(allocMain, tempSizes[2]); - if (!tempBuf[2] && tempSizes[2] != 0) - return SZ_ERROR_MEM; - - RINOK(LookInStream_SeekTo(inStream, startPos + offset)); - RINOK(SzDecodeCopy(s3Size, inStream, tempBuf[2])); - - if ((tempSizes[0] & 3) != 0 || - (tempSizes[1] & 3) != 0 || - tempSize3 + tempSizes[0] + tempSizes[1] != outSize) - return SZ_ERROR_DATA; - - { - CBcj2Dec p; - - p.bufs[0] = tempBuf3; p.lims[0] = tempBuf3 + tempSize3; - p.bufs[1] = tempBuf[0]; p.lims[1] = tempBuf[0] + tempSizes[0]; - p.bufs[2] = tempBuf[1]; p.lims[2] = tempBuf[1] + tempSizes[1]; - p.bufs[3] = tempBuf[2]; p.lims[3] = tempBuf[2] + tempSizes[2]; - - p.dest = outBuffer; - p.destLim = outBuffer + outSize; - - Bcj2Dec_Init(&p); - RINOK(Bcj2Dec_Decode(&p)); - - { - unsigned i; - for (i = 0; i < 4; i++) - if (p.bufs[i] != p.lims[i]) - return SZ_ERROR_DATA; - - if (!Bcj2Dec_IsFinished(&p)) - return SZ_ERROR_DATA; - - if (p.dest != p.destLim - || p.state != BCJ2_STREAM_MAIN) - return SZ_ERROR_DATA; - } - } - } - #ifndef _7Z_NO_METHODS_FILTERS - else if (ci == 1) - { - if (coder->MethodID == k_Delta) - { - if (coder->PropsSize != 1) - return SZ_ERROR_UNSUPPORTED; - { - Byte state[DELTA_STATE_SIZE]; - Delta_Init(state); - Delta_Decode(state, (unsigned)(propsData[coder->PropsOffset]) + 1, outBuffer, outSize); - } - } - else - { - if (coder->PropsSize != 0) - return SZ_ERROR_UNSUPPORTED; - switch (coder->MethodID) - { - case k_BCJ: - { - UInt32 state; - x86_Convert_Init(state); - x86_Convert(outBuffer, outSize, 0, &state, 0); - break; - } - CASE_BRA_CONV(PPC) - CASE_BRA_CONV(IA64) - CASE_BRA_CONV(SPARC) - CASE_BRA_CONV(ARM) - CASE_BRA_CONV(ARMT) - default: - return SZ_ERROR_UNSUPPORTED; - } - } - } - #endif - else - return SZ_ERROR_UNSUPPORTED; - } - - return SZ_OK; -} - - -SRes SzAr_DecodeFolder(const CSzAr *p, UInt32 folderIndex, - ILookInStream *inStream, UInt64 startPos, - Byte *outBuffer, size_t outSize, - ISzAllocPtr allocMain) -{ - SRes res; - CSzFolder folder; - CSzData sd; - - const Byte *data = p->CodersData + p->FoCodersOffsets[folderIndex]; - sd.Data = data; - sd.Size = p->FoCodersOffsets[(size_t)folderIndex + 1] - p->FoCodersOffsets[folderIndex]; - - res = SzGetNextFolderItem(&folder, &sd); - - if (res != SZ_OK) - return res; - - if (sd.Size != 0 - || folder.UnpackStream != p->FoToMainUnpackSizeIndex[folderIndex] - || outSize != SzAr_GetFolderUnpackSize(p, folderIndex)) - return SZ_ERROR_FAIL; - { - unsigned i; - Byte *tempBuf[3] = { 0, 0, 0}; - - res = SzFolder_Decode2(&folder, data, - &p->CoderUnpackSizes[p->FoToCoderUnpackSizes[folderIndex]], - p->PackPositions + p->FoStartPackStreamIndex[folderIndex], - inStream, startPos, - outBuffer, (SizeT)outSize, allocMain, tempBuf); - - for (i = 0; i < 3; i++) - ISzAlloc_Free(allocMain, tempBuf[i]); - - if (res == SZ_OK) - if (SzBitWithVals_Check(&p->FolderCRCs, folderIndex)) - if (CrcCalc(outBuffer, outSize) != p->FolderCRCs.Vals[folderIndex]) - res = SZ_ERROR_CRC; - - return res; - } -} +/* 7zDec.c -- Decoding from 7z folder +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +/* #define _7ZIP_PPMD_SUPPPORT */ + +#include "7z.h" +#include "7zCrc.h" + +#include "Bcj2.h" +#include "Bra.h" +#include "CpuArch.h" +#include "Delta.h" +#include "LzmaDec.h" +#include "Lzma2Dec.h" +#ifdef _7ZIP_PPMD_SUPPPORT +#include "Ppmd7.h" +#endif + +#define k_Copy 0 +#ifndef _7Z_NO_METHOD_LZMA2 +#define k_LZMA2 0x21 +#endif +#define k_LZMA 0x30101 +#define k_BCJ2 0x303011B +#ifndef _7Z_NO_METHODS_FILTERS +#define k_Delta 3 +#define k_BCJ 0x3030103 +#define k_PPC 0x3030205 +#define k_IA64 0x3030401 +#define k_ARM 0x3030501 +#define k_ARMT 0x3030701 +#define k_SPARC 0x3030805 +#endif + +#ifdef _7ZIP_PPMD_SUPPPORT + +#define k_PPMD 0x30401 + +typedef struct +{ + IByteIn vt; + const Byte *cur; + const Byte *end; + const Byte *begin; + UInt64 processed; + BoolInt extra; + SRes res; + const ILookInStream *inStream; +} CByteInToLook; + +static Byte ReadByte(const IByteIn *pp) +{ + CByteInToLook *p = CONTAINER_FROM_VTBL(pp, CByteInToLook, vt); + if (p->cur != p->end) + return *p->cur++; + if (p->res == SZ_OK) + { + size_t size = (size_t)(p->cur - p->begin); + p->processed += size; + p->res = ILookInStream_Skip(p->inStream, size); + size = (1 << 25); + p->res = ILookInStream_Look(p->inStream, (const void **)&p->begin, &size); + p->cur = p->begin; + p->end = p->begin + size; + if (size != 0) + return *p->cur++;; + } + p->extra = True; + return 0; +} + +static SRes SzDecodePpmd(const Byte *props, unsigned propsSize, UInt64 inSize, const ILookInStream *inStream, + Byte *outBuffer, SizeT outSize, ISzAllocPtr allocMain) +{ + CPpmd7 ppmd; + CByteInToLook s; + SRes res = SZ_OK; + + s.vt.Read = ReadByte; + s.inStream = inStream; + s.begin = s.end = s.cur = NULL; + s.extra = False; + s.res = SZ_OK; + s.processed = 0; + + if (propsSize != 5) + return SZ_ERROR_UNSUPPORTED; + + { + unsigned order = props[0]; + UInt32 memSize = GetUi32(props + 1); + if (order < PPMD7_MIN_ORDER || + order > PPMD7_MAX_ORDER || + memSize < PPMD7_MIN_MEM_SIZE || + memSize > PPMD7_MAX_MEM_SIZE) + return SZ_ERROR_UNSUPPORTED; + Ppmd7_Construct(&ppmd); + if (!Ppmd7_Alloc(&ppmd, memSize, allocMain)) + return SZ_ERROR_MEM; + Ppmd7_Init(&ppmd, order); + } + { + ppmd.rc.dec.Stream = &s.vt; + if (!Ppmd7z_RangeDec_Init(&ppmd.rc.dec)) + res = SZ_ERROR_DATA; + else if (!s.extra) + { + Byte *buf = outBuffer; + const Byte *lim = buf + outSize; + for (; buf != lim; buf++) + { + int sym = Ppmd7z_DecodeSymbol(&ppmd); + if (s.extra || sym < 0) + break; + *buf = (Byte)sym; + } + if (buf != lim) + res = SZ_ERROR_DATA; + else if (!Ppmd7z_RangeDec_IsFinishedOK(&ppmd.rc.dec)) + { + /* if (Ppmd7z_DecodeSymbol(&ppmd) != PPMD7_SYM_END || !Ppmd7z_RangeDec_IsFinishedOK(&ppmd.rc.dec)) */ + res = SZ_ERROR_DATA; + } + } + if (s.extra) + res = (s.res != SZ_OK ? s.res : SZ_ERROR_DATA); + else if (s.processed + (size_t)(s.cur - s.begin) != inSize) + res = SZ_ERROR_DATA; + } + Ppmd7_Free(&ppmd, allocMain); + return res; +} + +#endif + + +static SRes SzDecodeLzma(const Byte *props, unsigned propsSize, UInt64 inSize, ILookInStream *inStream, + Byte *outBuffer, SizeT outSize, ISzAllocPtr allocMain) +{ + CLzmaDec state; + SRes res = SZ_OK; + + LzmaDec_Construct(&state); + RINOK(LzmaDec_AllocateProbs(&state, props, propsSize, allocMain)); + state.dic = outBuffer; + state.dicBufSize = outSize; + LzmaDec_Init(&state); + + for (;;) + { + const void *inBuf = NULL; + size_t lookahead = (1 << 18); + if (lookahead > inSize) + lookahead = (size_t)inSize; + res = ILookInStream_Look(inStream, &inBuf, &lookahead); + if (res != SZ_OK) + break; + + { + SizeT inProcessed = (SizeT)lookahead, dicPos = state.dicPos; + ELzmaStatus status; + res = LzmaDec_DecodeToDic(&state, outSize, (const Byte *)inBuf, &inProcessed, LZMA_FINISH_END, &status); + lookahead -= inProcessed; + inSize -= inProcessed; + if (res != SZ_OK) + break; + + if (status == LZMA_STATUS_FINISHED_WITH_MARK) + { + if (outSize != state.dicPos || inSize != 0) + res = SZ_ERROR_DATA; + break; + } + + if (outSize == state.dicPos && inSize == 0 && status == LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK) + break; + + if (inProcessed == 0 && dicPos == state.dicPos) + { + res = SZ_ERROR_DATA; + break; + } + + res = ILookInStream_Skip(inStream, inProcessed); + if (res != SZ_OK) + break; + } + } + + LzmaDec_FreeProbs(&state, allocMain); + return res; +} + + +#ifndef _7Z_NO_METHOD_LZMA2 + +static SRes SzDecodeLzma2(const Byte *props, unsigned propsSize, UInt64 inSize, ILookInStream *inStream, + Byte *outBuffer, SizeT outSize, ISzAllocPtr allocMain) +{ + CLzma2Dec state; + SRes res = SZ_OK; + + Lzma2Dec_Construct(&state); + if (propsSize != 1) + return SZ_ERROR_DATA; + RINOK(Lzma2Dec_AllocateProbs(&state, props[0], allocMain)); + state.decoder.dic = outBuffer; + state.decoder.dicBufSize = outSize; + Lzma2Dec_Init(&state); + + for (;;) + { + const void *inBuf = NULL; + size_t lookahead = (1 << 18); + if (lookahead > inSize) + lookahead = (size_t)inSize; + res = ILookInStream_Look(inStream, &inBuf, &lookahead); + if (res != SZ_OK) + break; + + { + SizeT inProcessed = (SizeT)lookahead, dicPos = state.decoder.dicPos; + ELzmaStatus status; + res = Lzma2Dec_DecodeToDic(&state, outSize, (const Byte *)inBuf, &inProcessed, LZMA_FINISH_END, &status); + lookahead -= inProcessed; + inSize -= inProcessed; + if (res != SZ_OK) + break; + + if (status == LZMA_STATUS_FINISHED_WITH_MARK) + { + if (outSize != state.decoder.dicPos || inSize != 0) + res = SZ_ERROR_DATA; + break; + } + + if (inProcessed == 0 && dicPos == state.decoder.dicPos) + { + res = SZ_ERROR_DATA; + break; + } + + res = ILookInStream_Skip(inStream, inProcessed); + if (res != SZ_OK) + break; + } + } + + Lzma2Dec_FreeProbs(&state, allocMain); + return res; +} + +#endif + + +static SRes SzDecodeCopy(UInt64 inSize, ILookInStream *inStream, Byte *outBuffer) +{ + while (inSize > 0) + { + const void *inBuf; + size_t curSize = (1 << 18); + if (curSize > inSize) + curSize = (size_t)inSize; + RINOK(ILookInStream_Look(inStream, &inBuf, &curSize)); + if (curSize == 0) + return SZ_ERROR_INPUT_EOF; + memcpy(outBuffer, inBuf, curSize); + outBuffer += curSize; + inSize -= curSize; + RINOK(ILookInStream_Skip(inStream, curSize)); + } + return SZ_OK; +} + +static BoolInt IS_MAIN_METHOD(UInt32 m) +{ + switch (m) + { + case k_Copy: + case k_LZMA: + #ifndef _7Z_NO_METHOD_LZMA2 + case k_LZMA2: + #endif + #ifdef _7ZIP_PPMD_SUPPPORT + case k_PPMD: + #endif + return True; + } + return False; +} + +static BoolInt IS_SUPPORTED_CODER(const CSzCoderInfo *c) +{ + return + c->NumStreams == 1 + /* && c->MethodID <= (UInt32)0xFFFFFFFF */ + && IS_MAIN_METHOD((UInt32)c->MethodID); +} + +#define IS_BCJ2(c) ((c)->MethodID == k_BCJ2 && (c)->NumStreams == 4) + +static SRes CheckSupportedFolder(const CSzFolder *f) +{ + if (f->NumCoders < 1 || f->NumCoders > 4) + return SZ_ERROR_UNSUPPORTED; + if (!IS_SUPPORTED_CODER(&f->Coders[0])) + return SZ_ERROR_UNSUPPORTED; + if (f->NumCoders == 1) + { + if (f->NumPackStreams != 1 || f->PackStreams[0] != 0 || f->NumBonds != 0) + return SZ_ERROR_UNSUPPORTED; + return SZ_OK; + } + + + #ifndef _7Z_NO_METHODS_FILTERS + + if (f->NumCoders == 2) + { + const CSzCoderInfo *c = &f->Coders[1]; + if ( + /* c->MethodID > (UInt32)0xFFFFFFFF || */ + c->NumStreams != 1 + || f->NumPackStreams != 1 + || f->PackStreams[0] != 0 + || f->NumBonds != 1 + || f->Bonds[0].InIndex != 1 + || f->Bonds[0].OutIndex != 0) + return SZ_ERROR_UNSUPPORTED; + switch ((UInt32)c->MethodID) + { + case k_Delta: + case k_BCJ: + case k_PPC: + case k_IA64: + case k_SPARC: + case k_ARM: + case k_ARMT: + break; + default: + return SZ_ERROR_UNSUPPORTED; + } + return SZ_OK; + } + + #endif + + + if (f->NumCoders == 4) + { + if (!IS_SUPPORTED_CODER(&f->Coders[1]) + || !IS_SUPPORTED_CODER(&f->Coders[2]) + || !IS_BCJ2(&f->Coders[3])) + return SZ_ERROR_UNSUPPORTED; + if (f->NumPackStreams != 4 + || f->PackStreams[0] != 2 + || f->PackStreams[1] != 6 + || f->PackStreams[2] != 1 + || f->PackStreams[3] != 0 + || f->NumBonds != 3 + || f->Bonds[0].InIndex != 5 || f->Bonds[0].OutIndex != 0 + || f->Bonds[1].InIndex != 4 || f->Bonds[1].OutIndex != 1 + || f->Bonds[2].InIndex != 3 || f->Bonds[2].OutIndex != 2) + return SZ_ERROR_UNSUPPORTED; + return SZ_OK; + } + + return SZ_ERROR_UNSUPPORTED; +} + +#ifndef _7Z_NO_METHODS_FILTERS +#define CASE_BRA_CONV(isa) case k_ ## isa: isa ## _Convert(outBuffer, outSize, 0, 0); break; +#endif + +static SRes SzFolder_Decode2(const CSzFolder *folder, + const Byte *propsData, + const UInt64 *unpackSizes, + const UInt64 *packPositions, + ILookInStream *inStream, UInt64 startPos, + Byte *outBuffer, SizeT outSize, ISzAllocPtr allocMain, + Byte *tempBuf[]) +{ + UInt32 ci; + SizeT tempSizes[3] = { 0, 0, 0}; + SizeT tempSize3 = 0; + Byte *tempBuf3 = 0; + + RINOK(CheckSupportedFolder(folder)); + + for (ci = 0; ci < folder->NumCoders; ci++) + { + const CSzCoderInfo *coder = &folder->Coders[ci]; + + if (IS_MAIN_METHOD((UInt32)coder->MethodID)) + { + UInt32 si = 0; + UInt64 offset; + UInt64 inSize; + Byte *outBufCur = outBuffer; + SizeT outSizeCur = outSize; + if (folder->NumCoders == 4) + { + UInt32 indices[] = { 3, 2, 0 }; + UInt64 unpackSize = unpackSizes[ci]; + si = indices[ci]; + if (ci < 2) + { + Byte *temp; + outSizeCur = (SizeT)unpackSize; + if (outSizeCur != unpackSize) + return SZ_ERROR_MEM; + temp = (Byte *)ISzAlloc_Alloc(allocMain, outSizeCur); + if (!temp && outSizeCur != 0) + return SZ_ERROR_MEM; + outBufCur = tempBuf[1 - ci] = temp; + tempSizes[1 - ci] = outSizeCur; + } + else if (ci == 2) + { + if (unpackSize > outSize) /* check it */ + return SZ_ERROR_PARAM; + tempBuf3 = outBufCur = outBuffer + (outSize - (size_t)unpackSize); + tempSize3 = outSizeCur = (SizeT)unpackSize; + } + else + return SZ_ERROR_UNSUPPORTED; + } + offset = packPositions[si]; + inSize = packPositions[(size_t)si + 1] - offset; + RINOK(LookInStream_SeekTo(inStream, startPos + offset)); + + if (coder->MethodID == k_Copy) + { + if (inSize != outSizeCur) /* check it */ + return SZ_ERROR_DATA; + RINOK(SzDecodeCopy(inSize, inStream, outBufCur)); + } + else if (coder->MethodID == k_LZMA) + { + RINOK(SzDecodeLzma(propsData + coder->PropsOffset, coder->PropsSize, inSize, inStream, outBufCur, outSizeCur, allocMain)); + } + #ifndef _7Z_NO_METHOD_LZMA2 + else if (coder->MethodID == k_LZMA2) + { + RINOK(SzDecodeLzma2(propsData + coder->PropsOffset, coder->PropsSize, inSize, inStream, outBufCur, outSizeCur, allocMain)); + } + #endif + #ifdef _7ZIP_PPMD_SUPPPORT + else if (coder->MethodID == k_PPMD) + { + RINOK(SzDecodePpmd(propsData + coder->PropsOffset, coder->PropsSize, inSize, inStream, outBufCur, outSizeCur, allocMain)); + } + #endif + else + return SZ_ERROR_UNSUPPORTED; + } + else if (coder->MethodID == k_BCJ2) + { + UInt64 offset = packPositions[1]; + UInt64 s3Size = packPositions[2] - offset; + + if (ci != 3) + return SZ_ERROR_UNSUPPORTED; + + tempSizes[2] = (SizeT)s3Size; + if (tempSizes[2] != s3Size) + return SZ_ERROR_MEM; + tempBuf[2] = (Byte *)ISzAlloc_Alloc(allocMain, tempSizes[2]); + if (!tempBuf[2] && tempSizes[2] != 0) + return SZ_ERROR_MEM; + + RINOK(LookInStream_SeekTo(inStream, startPos + offset)); + RINOK(SzDecodeCopy(s3Size, inStream, tempBuf[2])); + + if ((tempSizes[0] & 3) != 0 || + (tempSizes[1] & 3) != 0 || + tempSize3 + tempSizes[0] + tempSizes[1] != outSize) + return SZ_ERROR_DATA; + + { + CBcj2Dec p; + + p.bufs[0] = tempBuf3; p.lims[0] = tempBuf3 + tempSize3; + p.bufs[1] = tempBuf[0]; p.lims[1] = tempBuf[0] + tempSizes[0]; + p.bufs[2] = tempBuf[1]; p.lims[2] = tempBuf[1] + tempSizes[1]; + p.bufs[3] = tempBuf[2]; p.lims[3] = tempBuf[2] + tempSizes[2]; + + p.dest = outBuffer; + p.destLim = outBuffer + outSize; + + Bcj2Dec_Init(&p); + RINOK(Bcj2Dec_Decode(&p)); + + { + unsigned i; + for (i = 0; i < 4; i++) + if (p.bufs[i] != p.lims[i]) + return SZ_ERROR_DATA; + + if (!Bcj2Dec_IsFinished(&p)) + return SZ_ERROR_DATA; + + if (p.dest != p.destLim + || p.state != BCJ2_STREAM_MAIN) + return SZ_ERROR_DATA; + } + } + } + #ifndef _7Z_NO_METHODS_FILTERS + else if (ci == 1) + { + if (coder->MethodID == k_Delta) + { + if (coder->PropsSize != 1) + return SZ_ERROR_UNSUPPORTED; + { + Byte state[DELTA_STATE_SIZE]; + Delta_Init(state); + Delta_Decode(state, (unsigned)(propsData[coder->PropsOffset]) + 1, outBuffer, outSize); + } + } + else + { + if (coder->PropsSize != 0) + return SZ_ERROR_UNSUPPORTED; + switch (coder->MethodID) + { + case k_BCJ: + { + UInt32 state; + x86_Convert_Init(state); + x86_Convert(outBuffer, outSize, 0, &state, 0); + break; + } + CASE_BRA_CONV(PPC) + CASE_BRA_CONV(IA64) + CASE_BRA_CONV(SPARC) + CASE_BRA_CONV(ARM) + CASE_BRA_CONV(ARMT) + default: + return SZ_ERROR_UNSUPPORTED; + } + } + } + #endif + else + return SZ_ERROR_UNSUPPORTED; + } + + return SZ_OK; +} + + +SRes SzAr_DecodeFolder(const CSzAr *p, UInt32 folderIndex, + ILookInStream *inStream, UInt64 startPos, + Byte *outBuffer, size_t outSize, + ISzAllocPtr allocMain) +{ + SRes res; + CSzFolder folder; + CSzData sd; + + const Byte *data = p->CodersData + p->FoCodersOffsets[folderIndex]; + sd.Data = data; + sd.Size = p->FoCodersOffsets[(size_t)folderIndex + 1] - p->FoCodersOffsets[folderIndex]; + + res = SzGetNextFolderItem(&folder, &sd); + + if (res != SZ_OK) + return res; + + if (sd.Size != 0 + || folder.UnpackStream != p->FoToMainUnpackSizeIndex[folderIndex] + || outSize != SzAr_GetFolderUnpackSize(p, folderIndex)) + return SZ_ERROR_FAIL; + { + unsigned i; + Byte *tempBuf[3] = { 0, 0, 0}; + + res = SzFolder_Decode2(&folder, data, + &p->CoderUnpackSizes[p->FoToCoderUnpackSizes[folderIndex]], + p->PackPositions + p->FoStartPackStreamIndex[folderIndex], + inStream, startPos, + outBuffer, (SizeT)outSize, allocMain, tempBuf); + + for (i = 0; i < 3; i++) + ISzAlloc_Free(allocMain, tempBuf[i]); + + if (res == SZ_OK) + if (SzBitWithVals_Check(&p->FolderCRCs, folderIndex)) + if (CrcCalc(outBuffer, outSize) != p->FolderCRCs.Vals[folderIndex]) + res = SZ_ERROR_CRC; + + return res; + } +} diff --git a/deps/LZMA-SDK/C/7zFile.c b/deps/LZMA-SDK/C/7zFile.c index 900125d52..13d2efa47 100644 --- a/deps/LZMA-SDK/C/7zFile.c +++ b/deps/LZMA-SDK/C/7zFile.c @@ -1,442 +1,442 @@ -/* 7zFile.c -- File IO -2021-04-29 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "7zFile.h" - -#ifndef USE_WINDOWS_FILE - - #include - - #ifndef USE_FOPEN - #include - #include - #ifdef _WIN32 - #include - typedef int ssize_t; - typedef int off_t; - #else - #include - #endif - #endif - -#else - -/* - ReadFile and WriteFile functions in Windows have BUG: - If you Read or Write 64MB or more (probably min_failure_size = 64MB - 32KB + 1) - from/to Network file, it returns ERROR_NO_SYSTEM_RESOURCES - (Insufficient system resources exist to complete the requested service). - Probably in some version of Windows there are problems with other sizes: - for 32 MB (maybe also for 16 MB). - And message can be "Network connection was lost" -*/ - -#endif - -#define kChunkSizeMax (1 << 22) - -void File_Construct(CSzFile *p) -{ - #ifdef USE_WINDOWS_FILE - p->handle = INVALID_HANDLE_VALUE; - #elif defined(USE_FOPEN) - p->file = NULL; - #else - p->fd = -1; - #endif -} - -#if !defined(UNDER_CE) || !defined(USE_WINDOWS_FILE) - -static WRes File_Open(CSzFile *p, const char *name, int writeMode) -{ - #ifdef USE_WINDOWS_FILE - - p->handle = CreateFileA(name, - writeMode ? GENERIC_WRITE : GENERIC_READ, - FILE_SHARE_READ, NULL, - writeMode ? CREATE_ALWAYS : OPEN_EXISTING, - FILE_ATTRIBUTE_NORMAL, NULL); - return (p->handle != INVALID_HANDLE_VALUE) ? 0 : GetLastError(); - - #elif defined(USE_FOPEN) - - p->file = fopen(name, writeMode ? "wb+" : "rb"); - return (p->file != 0) ? 0 : - #ifdef UNDER_CE - 2; /* ENOENT */ - #else - errno; - #endif - - #else - - int flags = (writeMode ? (O_CREAT | O_EXCL | O_WRONLY) : O_RDONLY); - #ifdef O_BINARY - flags |= O_BINARY; - #endif - p->fd = open(name, flags, 0666); - return (p->fd != -1) ? 0 : errno; - - #endif -} - -WRes InFile_Open(CSzFile *p, const char *name) { return File_Open(p, name, 0); } - -WRes OutFile_Open(CSzFile *p, const char *name) -{ - #if defined(USE_WINDOWS_FILE) || defined(USE_FOPEN) - return File_Open(p, name, 1); - #else - p->fd = creat(name, 0666); - return (p->fd != -1) ? 0 : errno; - #endif -} - -#endif - - -#ifdef USE_WINDOWS_FILE -static WRes File_OpenW(CSzFile *p, const WCHAR *name, int writeMode) -{ - p->handle = CreateFileW(name, - writeMode ? GENERIC_WRITE : GENERIC_READ, - FILE_SHARE_READ, NULL, - writeMode ? CREATE_ALWAYS : OPEN_EXISTING, - FILE_ATTRIBUTE_NORMAL, NULL); - return (p->handle != INVALID_HANDLE_VALUE) ? 0 : GetLastError(); -} -WRes InFile_OpenW(CSzFile *p, const WCHAR *name) { return File_OpenW(p, name, 0); } -WRes OutFile_OpenW(CSzFile *p, const WCHAR *name) { return File_OpenW(p, name, 1); } -#endif - -WRes File_Close(CSzFile *p) -{ - #ifdef USE_WINDOWS_FILE - - if (p->handle != INVALID_HANDLE_VALUE) - { - if (!CloseHandle(p->handle)) - return GetLastError(); - p->handle = INVALID_HANDLE_VALUE; - } - - #elif defined(USE_FOPEN) - - if (p->file != NULL) - { - int res = fclose(p->file); - if (res != 0) - { - if (res == EOF) - return errno; - return res; - } - p->file = NULL; - } - - #else - - if (p->fd != -1) - { - if (close(p->fd) != 0) - return errno; - p->fd = -1; - } - - #endif - - return 0; -} - - -WRes File_Read(CSzFile *p, void *data, size_t *size) -{ - size_t originalSize = *size; - *size = 0; - if (originalSize == 0) - return 0; - - #ifdef USE_WINDOWS_FILE - - do - { - const DWORD curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : (DWORD)originalSize; - DWORD processed = 0; - const BOOL res = ReadFile(p->handle, data, curSize, &processed, NULL); - data = (void *)((Byte *)data + processed); - originalSize -= processed; - *size += processed; - if (!res) - return GetLastError(); - // debug : we can break here for partial reading mode - if (processed == 0) - break; - } - while (originalSize > 0); - - #elif defined(USE_FOPEN) - - do - { - const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize; - const size_t processed = fread(data, 1, curSize, p->file); - data = (void *)((Byte *)data + (size_t)processed); - originalSize -= processed; - *size += processed; - if (processed != curSize) - return ferror(p->file); - // debug : we can break here for partial reading mode - if (processed == 0) - break; - } - while (originalSize > 0); - - #else - - do - { - const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize; - const ssize_t processed = read(p->fd, data, curSize); - if (processed == -1) - return errno; - if (processed == 0) - break; - data = (void *)((Byte *)data + (size_t)processed); - originalSize -= (size_t)processed; - *size += (size_t)processed; - // debug : we can break here for partial reading mode - // break; - } - while (originalSize > 0); - - #endif - - return 0; -} - - -WRes File_Write(CSzFile *p, const void *data, size_t *size) -{ - size_t originalSize = *size; - *size = 0; - if (originalSize == 0) - return 0; - - #ifdef USE_WINDOWS_FILE - - do - { - const DWORD curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : (DWORD)originalSize; - DWORD processed = 0; - const BOOL res = WriteFile(p->handle, data, curSize, &processed, NULL); - data = (const void *)((const Byte *)data + processed); - originalSize -= processed; - *size += processed; - if (!res) - return GetLastError(); - if (processed == 0) - break; - } - while (originalSize > 0); - - #elif defined(USE_FOPEN) - - do - { - const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize; - const size_t processed = fwrite(data, 1, curSize, p->file); - data = (void *)((Byte *)data + (size_t)processed); - originalSize -= processed; - *size += processed; - if (processed != curSize) - return ferror(p->file); - if (processed == 0) - break; - } - while (originalSize > 0); - - #else - - do - { - const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize; - const ssize_t processed = write(p->fd, data, curSize); - if (processed == -1) - return errno; - if (processed == 0) - break; - data = (void *)((Byte *)data + (size_t)processed); - originalSize -= (size_t)processed; - *size += (size_t)processed; - } - while (originalSize > 0); - - #endif - - return 0; -} - - -WRes File_Seek(CSzFile *p, Int64 *pos, ESzSeek origin) -{ - #ifdef USE_WINDOWS_FILE - - DWORD moveMethod; - UInt32 low = (UInt32)*pos; - LONG high = (LONG)((UInt64)*pos >> 16 >> 16); /* for case when UInt64 is 32-bit only */ - switch (origin) - { - case SZ_SEEK_SET: moveMethod = FILE_BEGIN; break; - case SZ_SEEK_CUR: moveMethod = FILE_CURRENT; break; - case SZ_SEEK_END: moveMethod = FILE_END; break; - default: return ERROR_INVALID_PARAMETER; - } - low = SetFilePointer(p->handle, (LONG)low, &high, moveMethod); - if (low == (UInt32)0xFFFFFFFF) - { - WRes res = GetLastError(); - if (res != NO_ERROR) - return res; - } - *pos = ((Int64)high << 32) | low; - return 0; - - #else - - int moveMethod; // = origin; - - switch (origin) - { - case SZ_SEEK_SET: moveMethod = SEEK_SET; break; - case SZ_SEEK_CUR: moveMethod = SEEK_CUR; break; - case SZ_SEEK_END: moveMethod = SEEK_END; break; - default: return EINVAL; - } - - #if defined(USE_FOPEN) - { - int res = fseek(p->file, (long)*pos, moveMethod); - if (res == -1) - return errno; - *pos = ftell(p->file); - if (*pos == -1) - return errno; - return 0; - } - #else - { - off_t res = lseek(p->fd, (off_t)*pos, moveMethod); - if (res == -1) - return errno; - *pos = res; - return 0; - } - - #endif // USE_FOPEN - #endif // USE_WINDOWS_FILE -} - - -WRes File_GetLength(CSzFile *p, UInt64 *length) -{ - #ifdef USE_WINDOWS_FILE - - DWORD sizeHigh; - DWORD sizeLow = GetFileSize(p->handle, &sizeHigh); - if (sizeLow == 0xFFFFFFFF) - { - DWORD res = GetLastError(); - if (res != NO_ERROR) - return res; - } - *length = (((UInt64)sizeHigh) << 32) + sizeLow; - return 0; - - #elif defined(USE_FOPEN) - - long pos = ftell(p->file); - int res = fseek(p->file, 0, SEEK_END); - *length = ftell(p->file); - fseek(p->file, pos, SEEK_SET); - return res; - - #else - - off_t pos; - *length = 0; - pos = lseek(p->fd, 0, SEEK_CUR); - if (pos != -1) - { - const off_t len2 = lseek(p->fd, 0, SEEK_END); - const off_t res2 = lseek(p->fd, pos, SEEK_SET); - if (len2 != -1) - { - *length = (UInt64)len2; - if (res2 != -1) - return 0; - } - } - return errno; - - #endif -} - - -/* ---------- FileSeqInStream ---------- */ - -static SRes FileSeqInStream_Read(const ISeqInStream *pp, void *buf, size_t *size) -{ - CFileSeqInStream *p = CONTAINER_FROM_VTBL(pp, CFileSeqInStream, vt); - WRes wres = File_Read(&p->file, buf, size); - p->wres = wres; - return (wres == 0) ? SZ_OK : SZ_ERROR_READ; -} - -void FileSeqInStream_CreateVTable(CFileSeqInStream *p) -{ - p->vt.Read = FileSeqInStream_Read; -} - - -/* ---------- FileInStream ---------- */ - -static SRes FileInStream_Read(const ISeekInStream *pp, void *buf, size_t *size) -{ - CFileInStream *p = CONTAINER_FROM_VTBL(pp, CFileInStream, vt); - WRes wres = File_Read(&p->file, buf, size); - p->wres = wres; - return (wres == 0) ? SZ_OK : SZ_ERROR_READ; -} - -static SRes FileInStream_Seek(const ISeekInStream *pp, Int64 *pos, ESzSeek origin) -{ - CFileInStream *p = CONTAINER_FROM_VTBL(pp, CFileInStream, vt); - WRes wres = File_Seek(&p->file, pos, origin); - p->wres = wres; - return (wres == 0) ? SZ_OK : SZ_ERROR_READ; -} - -void FileInStream_CreateVTable(CFileInStream *p) -{ - p->vt.Read = FileInStream_Read; - p->vt.Seek = FileInStream_Seek; -} - - -/* ---------- FileOutStream ---------- */ - -static size_t FileOutStream_Write(const ISeqOutStream *pp, const void *data, size_t size) -{ - CFileOutStream *p = CONTAINER_FROM_VTBL(pp, CFileOutStream, vt); - WRes wres = File_Write(&p->file, data, &size); - p->wres = wres; - return size; -} - -void FileOutStream_CreateVTable(CFileOutStream *p) -{ - p->vt.Write = FileOutStream_Write; -} +/* 7zFile.c -- File IO +2021-04-29 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "7zFile.h" + +#ifndef USE_WINDOWS_FILE + + #include + + #ifndef USE_FOPEN + #include + #include + #ifdef _WIN32 + #include + typedef int ssize_t; + typedef int off_t; + #else + #include + #endif + #endif + +#else + +/* + ReadFile and WriteFile functions in Windows have BUG: + If you Read or Write 64MB or more (probably min_failure_size = 64MB - 32KB + 1) + from/to Network file, it returns ERROR_NO_SYSTEM_RESOURCES + (Insufficient system resources exist to complete the requested service). + Probably in some version of Windows there are problems with other sizes: + for 32 MB (maybe also for 16 MB). + And message can be "Network connection was lost" +*/ + +#endif + +#define kChunkSizeMax (1 << 22) + +void File_Construct(CSzFile *p) +{ + #ifdef USE_WINDOWS_FILE + p->handle = INVALID_HANDLE_VALUE; + #elif defined(USE_FOPEN) + p->file = NULL; + #else + p->fd = -1; + #endif +} + +#if !defined(UNDER_CE) || !defined(USE_WINDOWS_FILE) + +static WRes File_Open(CSzFile *p, const char *name, int writeMode) +{ + #ifdef USE_WINDOWS_FILE + + p->handle = CreateFileA(name, + writeMode ? GENERIC_WRITE : GENERIC_READ, + FILE_SHARE_READ, NULL, + writeMode ? CREATE_ALWAYS : OPEN_EXISTING, + FILE_ATTRIBUTE_NORMAL, NULL); + return (p->handle != INVALID_HANDLE_VALUE) ? 0 : GetLastError(); + + #elif defined(USE_FOPEN) + + p->file = fopen(name, writeMode ? "wb+" : "rb"); + return (p->file != 0) ? 0 : + #ifdef UNDER_CE + 2; /* ENOENT */ + #else + errno; + #endif + + #else + + int flags = (writeMode ? (O_CREAT | O_EXCL | O_WRONLY) : O_RDONLY); + #ifdef O_BINARY + flags |= O_BINARY; + #endif + p->fd = open(name, flags, 0666); + return (p->fd != -1) ? 0 : errno; + + #endif +} + +WRes InFile_Open(CSzFile *p, const char *name) { return File_Open(p, name, 0); } + +WRes OutFile_Open(CSzFile *p, const char *name) +{ + #if defined(USE_WINDOWS_FILE) || defined(USE_FOPEN) + return File_Open(p, name, 1); + #else + p->fd = creat(name, 0666); + return (p->fd != -1) ? 0 : errno; + #endif +} + +#endif + + +#ifdef USE_WINDOWS_FILE +static WRes File_OpenW(CSzFile *p, const WCHAR *name, int writeMode) +{ + p->handle = CreateFileW(name, + writeMode ? GENERIC_WRITE : GENERIC_READ, + FILE_SHARE_READ, NULL, + writeMode ? CREATE_ALWAYS : OPEN_EXISTING, + FILE_ATTRIBUTE_NORMAL, NULL); + return (p->handle != INVALID_HANDLE_VALUE) ? 0 : GetLastError(); +} +WRes InFile_OpenW(CSzFile *p, const WCHAR *name) { return File_OpenW(p, name, 0); } +WRes OutFile_OpenW(CSzFile *p, const WCHAR *name) { return File_OpenW(p, name, 1); } +#endif + +WRes File_Close(CSzFile *p) +{ + #ifdef USE_WINDOWS_FILE + + if (p->handle != INVALID_HANDLE_VALUE) + { + if (!CloseHandle(p->handle)) + return GetLastError(); + p->handle = INVALID_HANDLE_VALUE; + } + + #elif defined(USE_FOPEN) + + if (p->file != NULL) + { + int res = fclose(p->file); + if (res != 0) + { + if (res == EOF) + return errno; + return res; + } + p->file = NULL; + } + + #else + + if (p->fd != -1) + { + if (close(p->fd) != 0) + return errno; + p->fd = -1; + } + + #endif + + return 0; +} + + +WRes File_Read(CSzFile *p, void *data, size_t *size) +{ + size_t originalSize = *size; + *size = 0; + if (originalSize == 0) + return 0; + + #ifdef USE_WINDOWS_FILE + + do + { + const DWORD curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : (DWORD)originalSize; + DWORD processed = 0; + const BOOL res = ReadFile(p->handle, data, curSize, &processed, NULL); + data = (void *)((Byte *)data + processed); + originalSize -= processed; + *size += processed; + if (!res) + return GetLastError(); + // debug : we can break here for partial reading mode + if (processed == 0) + break; + } + while (originalSize > 0); + + #elif defined(USE_FOPEN) + + do + { + const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize; + const size_t processed = fread(data, 1, curSize, p->file); + data = (void *)((Byte *)data + (size_t)processed); + originalSize -= processed; + *size += processed; + if (processed != curSize) + return ferror(p->file); + // debug : we can break here for partial reading mode + if (processed == 0) + break; + } + while (originalSize > 0); + + #else + + do + { + const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize; + const ssize_t processed = read(p->fd, data, curSize); + if (processed == -1) + return errno; + if (processed == 0) + break; + data = (void *)((Byte *)data + (size_t)processed); + originalSize -= (size_t)processed; + *size += (size_t)processed; + // debug : we can break here for partial reading mode + // break; + } + while (originalSize > 0); + + #endif + + return 0; +} + + +WRes File_Write(CSzFile *p, const void *data, size_t *size) +{ + size_t originalSize = *size; + *size = 0; + if (originalSize == 0) + return 0; + + #ifdef USE_WINDOWS_FILE + + do + { + const DWORD curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : (DWORD)originalSize; + DWORD processed = 0; + const BOOL res = WriteFile(p->handle, data, curSize, &processed, NULL); + data = (const void *)((const Byte *)data + processed); + originalSize -= processed; + *size += processed; + if (!res) + return GetLastError(); + if (processed == 0) + break; + } + while (originalSize > 0); + + #elif defined(USE_FOPEN) + + do + { + const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize; + const size_t processed = fwrite(data, 1, curSize, p->file); + data = (void *)((Byte *)data + (size_t)processed); + originalSize -= processed; + *size += processed; + if (processed != curSize) + return ferror(p->file); + if (processed == 0) + break; + } + while (originalSize > 0); + + #else + + do + { + const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize; + const ssize_t processed = write(p->fd, data, curSize); + if (processed == -1) + return errno; + if (processed == 0) + break; + data = (void *)((Byte *)data + (size_t)processed); + originalSize -= (size_t)processed; + *size += (size_t)processed; + } + while (originalSize > 0); + + #endif + + return 0; +} + + +WRes File_Seek(CSzFile *p, Int64 *pos, ESzSeek origin) +{ + #ifdef USE_WINDOWS_FILE + + DWORD moveMethod; + UInt32 low = (UInt32)*pos; + LONG high = (LONG)((UInt64)*pos >> 16 >> 16); /* for case when UInt64 is 32-bit only */ + switch (origin) + { + case SZ_SEEK_SET: moveMethod = FILE_BEGIN; break; + case SZ_SEEK_CUR: moveMethod = FILE_CURRENT; break; + case SZ_SEEK_END: moveMethod = FILE_END; break; + default: return ERROR_INVALID_PARAMETER; + } + low = SetFilePointer(p->handle, (LONG)low, &high, moveMethod); + if (low == (UInt32)0xFFFFFFFF) + { + WRes res = GetLastError(); + if (res != NO_ERROR) + return res; + } + *pos = ((Int64)high << 32) | low; + return 0; + + #else + + int moveMethod; // = origin; + + switch (origin) + { + case SZ_SEEK_SET: moveMethod = SEEK_SET; break; + case SZ_SEEK_CUR: moveMethod = SEEK_CUR; break; + case SZ_SEEK_END: moveMethod = SEEK_END; break; + default: return EINVAL; + } + + #if defined(USE_FOPEN) + { + int res = fseek(p->file, (long)*pos, moveMethod); + if (res == -1) + return errno; + *pos = ftell(p->file); + if (*pos == -1) + return errno; + return 0; + } + #else + { + off_t res = lseek(p->fd, (off_t)*pos, moveMethod); + if (res == -1) + return errno; + *pos = res; + return 0; + } + + #endif // USE_FOPEN + #endif // USE_WINDOWS_FILE +} + + +WRes File_GetLength(CSzFile *p, UInt64 *length) +{ + #ifdef USE_WINDOWS_FILE + + DWORD sizeHigh; + DWORD sizeLow = GetFileSize(p->handle, &sizeHigh); + if (sizeLow == 0xFFFFFFFF) + { + DWORD res = GetLastError(); + if (res != NO_ERROR) + return res; + } + *length = (((UInt64)sizeHigh) << 32) + sizeLow; + return 0; + + #elif defined(USE_FOPEN) + + long pos = ftell(p->file); + int res = fseek(p->file, 0, SEEK_END); + *length = ftell(p->file); + fseek(p->file, pos, SEEK_SET); + return res; + + #else + + off_t pos; + *length = 0; + pos = lseek(p->fd, 0, SEEK_CUR); + if (pos != -1) + { + const off_t len2 = lseek(p->fd, 0, SEEK_END); + const off_t res2 = lseek(p->fd, pos, SEEK_SET); + if (len2 != -1) + { + *length = (UInt64)len2; + if (res2 != -1) + return 0; + } + } + return errno; + + #endif +} + + +/* ---------- FileSeqInStream ---------- */ + +static SRes FileSeqInStream_Read(const ISeqInStream *pp, void *buf, size_t *size) +{ + CFileSeqInStream *p = CONTAINER_FROM_VTBL(pp, CFileSeqInStream, vt); + WRes wres = File_Read(&p->file, buf, size); + p->wres = wres; + return (wres == 0) ? SZ_OK : SZ_ERROR_READ; +} + +void FileSeqInStream_CreateVTable(CFileSeqInStream *p) +{ + p->vt.Read = FileSeqInStream_Read; +} + + +/* ---------- FileInStream ---------- */ + +static SRes FileInStream_Read(const ISeekInStream *pp, void *buf, size_t *size) +{ + CFileInStream *p = CONTAINER_FROM_VTBL(pp, CFileInStream, vt); + WRes wres = File_Read(&p->file, buf, size); + p->wres = wres; + return (wres == 0) ? SZ_OK : SZ_ERROR_READ; +} + +static SRes FileInStream_Seek(const ISeekInStream *pp, Int64 *pos, ESzSeek origin) +{ + CFileInStream *p = CONTAINER_FROM_VTBL(pp, CFileInStream, vt); + WRes wres = File_Seek(&p->file, pos, origin); + p->wres = wres; + return (wres == 0) ? SZ_OK : SZ_ERROR_READ; +} + +void FileInStream_CreateVTable(CFileInStream *p) +{ + p->vt.Read = FileInStream_Read; + p->vt.Seek = FileInStream_Seek; +} + + +/* ---------- FileOutStream ---------- */ + +static size_t FileOutStream_Write(const ISeqOutStream *pp, const void *data, size_t size) +{ + CFileOutStream *p = CONTAINER_FROM_VTBL(pp, CFileOutStream, vt); + WRes wres = File_Write(&p->file, data, &size); + p->wres = wres; + return size; +} + +void FileOutStream_CreateVTable(CFileOutStream *p) +{ + p->vt.Write = FileOutStream_Write; +} diff --git a/deps/LZMA-SDK/C/7zFile.h b/deps/LZMA-SDK/C/7zFile.h index c7a30fc2b..788abb6b9 100644 --- a/deps/LZMA-SDK/C/7zFile.h +++ b/deps/LZMA-SDK/C/7zFile.h @@ -1,91 +1,91 @@ -/* 7zFile.h -- File IO -2021-02-15 : Igor Pavlov : Public domain */ - -#ifndef __7Z_FILE_H -#define __7Z_FILE_H - -#ifdef _WIN32 -#define USE_WINDOWS_FILE -// #include -#endif - -#ifdef USE_WINDOWS_FILE -#include -#else -// note: USE_FOPEN mode is limited to 32-bit file size -// #define USE_FOPEN -// #include -#endif - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -/* ---------- File ---------- */ - -typedef struct -{ - #ifdef USE_WINDOWS_FILE - HANDLE handle; - #elif defined(USE_FOPEN) - FILE *file; - #else - int fd; - #endif -} CSzFile; - -void File_Construct(CSzFile *p); -#if !defined(UNDER_CE) || !defined(USE_WINDOWS_FILE) -WRes InFile_Open(CSzFile *p, const char *name); -WRes OutFile_Open(CSzFile *p, const char *name); -#endif -#ifdef USE_WINDOWS_FILE -WRes InFile_OpenW(CSzFile *p, const WCHAR *name); -WRes OutFile_OpenW(CSzFile *p, const WCHAR *name); -#endif -WRes File_Close(CSzFile *p); - -/* reads max(*size, remain file's size) bytes */ -WRes File_Read(CSzFile *p, void *data, size_t *size); - -/* writes *size bytes */ -WRes File_Write(CSzFile *p, const void *data, size_t *size); - -WRes File_Seek(CSzFile *p, Int64 *pos, ESzSeek origin); -WRes File_GetLength(CSzFile *p, UInt64 *length); - - -/* ---------- FileInStream ---------- */ - -typedef struct -{ - ISeqInStream vt; - CSzFile file; - WRes wres; -} CFileSeqInStream; - -void FileSeqInStream_CreateVTable(CFileSeqInStream *p); - - -typedef struct -{ - ISeekInStream vt; - CSzFile file; - WRes wres; -} CFileInStream; - -void FileInStream_CreateVTable(CFileInStream *p); - - -typedef struct -{ - ISeqOutStream vt; - CSzFile file; - WRes wres; -} CFileOutStream; - -void FileOutStream_CreateVTable(CFileOutStream *p); - -EXTERN_C_END - -#endif +/* 7zFile.h -- File IO +2021-02-15 : Igor Pavlov : Public domain */ + +#ifndef __7Z_FILE_H +#define __7Z_FILE_H + +#ifdef _WIN32 +#define USE_WINDOWS_FILE +// #include +#endif + +#ifdef USE_WINDOWS_FILE +#include +#else +// note: USE_FOPEN mode is limited to 32-bit file size +// #define USE_FOPEN +// #include +#endif + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +/* ---------- File ---------- */ + +typedef struct +{ + #ifdef USE_WINDOWS_FILE + HANDLE handle; + #elif defined(USE_FOPEN) + FILE *file; + #else + int fd; + #endif +} CSzFile; + +void File_Construct(CSzFile *p); +#if !defined(UNDER_CE) || !defined(USE_WINDOWS_FILE) +WRes InFile_Open(CSzFile *p, const char *name); +WRes OutFile_Open(CSzFile *p, const char *name); +#endif +#ifdef USE_WINDOWS_FILE +WRes InFile_OpenW(CSzFile *p, const WCHAR *name); +WRes OutFile_OpenW(CSzFile *p, const WCHAR *name); +#endif +WRes File_Close(CSzFile *p); + +/* reads max(*size, remain file's size) bytes */ +WRes File_Read(CSzFile *p, void *data, size_t *size); + +/* writes *size bytes */ +WRes File_Write(CSzFile *p, const void *data, size_t *size); + +WRes File_Seek(CSzFile *p, Int64 *pos, ESzSeek origin); +WRes File_GetLength(CSzFile *p, UInt64 *length); + + +/* ---------- FileInStream ---------- */ + +typedef struct +{ + ISeqInStream vt; + CSzFile file; + WRes wres; +} CFileSeqInStream; + +void FileSeqInStream_CreateVTable(CFileSeqInStream *p); + + +typedef struct +{ + ISeekInStream vt; + CSzFile file; + WRes wres; +} CFileInStream; + +void FileInStream_CreateVTable(CFileInStream *p); + + +typedef struct +{ + ISeqOutStream vt; + CSzFile file; + WRes wres; +} CFileOutStream; + +void FileOutStream_CreateVTable(CFileOutStream *p); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/7zStream.c b/deps/LZMA-SDK/C/7zStream.c index 4b472a41d..28a14604f 100644 --- a/deps/LZMA-SDK/C/7zStream.c +++ b/deps/LZMA-SDK/C/7zStream.c @@ -1,176 +1,176 @@ -/* 7zStream.c -- 7z Stream functions -2021-02-09 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include - -#include "7zTypes.h" - -SRes SeqInStream_Read2(const ISeqInStream *stream, void *buf, size_t size, SRes errorType) -{ - while (size != 0) - { - size_t processed = size; - RINOK(ISeqInStream_Read(stream, buf, &processed)); - if (processed == 0) - return errorType; - buf = (void *)((Byte *)buf + processed); - size -= processed; - } - return SZ_OK; -} - -SRes SeqInStream_Read(const ISeqInStream *stream, void *buf, size_t size) -{ - return SeqInStream_Read2(stream, buf, size, SZ_ERROR_INPUT_EOF); -} - -SRes SeqInStream_ReadByte(const ISeqInStream *stream, Byte *buf) -{ - size_t processed = 1; - RINOK(ISeqInStream_Read(stream, buf, &processed)); - return (processed == 1) ? SZ_OK : SZ_ERROR_INPUT_EOF; -} - - - -SRes LookInStream_SeekTo(const ILookInStream *stream, UInt64 offset) -{ - Int64 t = (Int64)offset; - return ILookInStream_Seek(stream, &t, SZ_SEEK_SET); -} - -SRes LookInStream_LookRead(const ILookInStream *stream, void *buf, size_t *size) -{ - const void *lookBuf; - if (*size == 0) - return SZ_OK; - RINOK(ILookInStream_Look(stream, &lookBuf, size)); - memcpy(buf, lookBuf, *size); - return ILookInStream_Skip(stream, *size); -} - -SRes LookInStream_Read2(const ILookInStream *stream, void *buf, size_t size, SRes errorType) -{ - while (size != 0) - { - size_t processed = size; - RINOK(ILookInStream_Read(stream, buf, &processed)); - if (processed == 0) - return errorType; - buf = (void *)((Byte *)buf + processed); - size -= processed; - } - return SZ_OK; -} - -SRes LookInStream_Read(const ILookInStream *stream, void *buf, size_t size) -{ - return LookInStream_Read2(stream, buf, size, SZ_ERROR_INPUT_EOF); -} - - - -#define GET_LookToRead2 CLookToRead2 *p = CONTAINER_FROM_VTBL(pp, CLookToRead2, vt); - -static SRes LookToRead2_Look_Lookahead(const ILookInStream *pp, const void **buf, size_t *size) -{ - SRes res = SZ_OK; - GET_LookToRead2 - size_t size2 = p->size - p->pos; - if (size2 == 0 && *size != 0) - { - p->pos = 0; - p->size = 0; - size2 = p->bufSize; - res = ISeekInStream_Read(p->realStream, p->buf, &size2); - p->size = size2; - } - if (*size > size2) - *size = size2; - *buf = p->buf + p->pos; - return res; -} - -static SRes LookToRead2_Look_Exact(const ILookInStream *pp, const void **buf, size_t *size) -{ - SRes res = SZ_OK; - GET_LookToRead2 - size_t size2 = p->size - p->pos; - if (size2 == 0 && *size != 0) - { - p->pos = 0; - p->size = 0; - if (*size > p->bufSize) - *size = p->bufSize; - res = ISeekInStream_Read(p->realStream, p->buf, size); - size2 = p->size = *size; - } - if (*size > size2) - *size = size2; - *buf = p->buf + p->pos; - return res; -} - -static SRes LookToRead2_Skip(const ILookInStream *pp, size_t offset) -{ - GET_LookToRead2 - p->pos += offset; - return SZ_OK; -} - -static SRes LookToRead2_Read(const ILookInStream *pp, void *buf, size_t *size) -{ - GET_LookToRead2 - size_t rem = p->size - p->pos; - if (rem == 0) - return ISeekInStream_Read(p->realStream, buf, size); - if (rem > *size) - rem = *size; - memcpy(buf, p->buf + p->pos, rem); - p->pos += rem; - *size = rem; - return SZ_OK; -} - -static SRes LookToRead2_Seek(const ILookInStream *pp, Int64 *pos, ESzSeek origin) -{ - GET_LookToRead2 - p->pos = p->size = 0; - return ISeekInStream_Seek(p->realStream, pos, origin); -} - -void LookToRead2_CreateVTable(CLookToRead2 *p, int lookahead) -{ - p->vt.Look = lookahead ? - LookToRead2_Look_Lookahead : - LookToRead2_Look_Exact; - p->vt.Skip = LookToRead2_Skip; - p->vt.Read = LookToRead2_Read; - p->vt.Seek = LookToRead2_Seek; -} - - - -static SRes SecToLook_Read(const ISeqInStream *pp, void *buf, size_t *size) -{ - CSecToLook *p = CONTAINER_FROM_VTBL(pp, CSecToLook, vt); - return LookInStream_LookRead(p->realStream, buf, size); -} - -void SecToLook_CreateVTable(CSecToLook *p) -{ - p->vt.Read = SecToLook_Read; -} - -static SRes SecToRead_Read(const ISeqInStream *pp, void *buf, size_t *size) -{ - CSecToRead *p = CONTAINER_FROM_VTBL(pp, CSecToRead, vt); - return ILookInStream_Read(p->realStream, buf, size); -} - -void SecToRead_CreateVTable(CSecToRead *p) -{ - p->vt.Read = SecToRead_Read; -} +/* 7zStream.c -- 7z Stream functions +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#include "7zTypes.h" + +SRes SeqInStream_Read2(const ISeqInStream *stream, void *buf, size_t size, SRes errorType) +{ + while (size != 0) + { + size_t processed = size; + RINOK(ISeqInStream_Read(stream, buf, &processed)); + if (processed == 0) + return errorType; + buf = (void *)((Byte *)buf + processed); + size -= processed; + } + return SZ_OK; +} + +SRes SeqInStream_Read(const ISeqInStream *stream, void *buf, size_t size) +{ + return SeqInStream_Read2(stream, buf, size, SZ_ERROR_INPUT_EOF); +} + +SRes SeqInStream_ReadByte(const ISeqInStream *stream, Byte *buf) +{ + size_t processed = 1; + RINOK(ISeqInStream_Read(stream, buf, &processed)); + return (processed == 1) ? SZ_OK : SZ_ERROR_INPUT_EOF; +} + + + +SRes LookInStream_SeekTo(const ILookInStream *stream, UInt64 offset) +{ + Int64 t = (Int64)offset; + return ILookInStream_Seek(stream, &t, SZ_SEEK_SET); +} + +SRes LookInStream_LookRead(const ILookInStream *stream, void *buf, size_t *size) +{ + const void *lookBuf; + if (*size == 0) + return SZ_OK; + RINOK(ILookInStream_Look(stream, &lookBuf, size)); + memcpy(buf, lookBuf, *size); + return ILookInStream_Skip(stream, *size); +} + +SRes LookInStream_Read2(const ILookInStream *stream, void *buf, size_t size, SRes errorType) +{ + while (size != 0) + { + size_t processed = size; + RINOK(ILookInStream_Read(stream, buf, &processed)); + if (processed == 0) + return errorType; + buf = (void *)((Byte *)buf + processed); + size -= processed; + } + return SZ_OK; +} + +SRes LookInStream_Read(const ILookInStream *stream, void *buf, size_t size) +{ + return LookInStream_Read2(stream, buf, size, SZ_ERROR_INPUT_EOF); +} + + + +#define GET_LookToRead2 CLookToRead2 *p = CONTAINER_FROM_VTBL(pp, CLookToRead2, vt); + +static SRes LookToRead2_Look_Lookahead(const ILookInStream *pp, const void **buf, size_t *size) +{ + SRes res = SZ_OK; + GET_LookToRead2 + size_t size2 = p->size - p->pos; + if (size2 == 0 && *size != 0) + { + p->pos = 0; + p->size = 0; + size2 = p->bufSize; + res = ISeekInStream_Read(p->realStream, p->buf, &size2); + p->size = size2; + } + if (*size > size2) + *size = size2; + *buf = p->buf + p->pos; + return res; +} + +static SRes LookToRead2_Look_Exact(const ILookInStream *pp, const void **buf, size_t *size) +{ + SRes res = SZ_OK; + GET_LookToRead2 + size_t size2 = p->size - p->pos; + if (size2 == 0 && *size != 0) + { + p->pos = 0; + p->size = 0; + if (*size > p->bufSize) + *size = p->bufSize; + res = ISeekInStream_Read(p->realStream, p->buf, size); + size2 = p->size = *size; + } + if (*size > size2) + *size = size2; + *buf = p->buf + p->pos; + return res; +} + +static SRes LookToRead2_Skip(const ILookInStream *pp, size_t offset) +{ + GET_LookToRead2 + p->pos += offset; + return SZ_OK; +} + +static SRes LookToRead2_Read(const ILookInStream *pp, void *buf, size_t *size) +{ + GET_LookToRead2 + size_t rem = p->size - p->pos; + if (rem == 0) + return ISeekInStream_Read(p->realStream, buf, size); + if (rem > *size) + rem = *size; + memcpy(buf, p->buf + p->pos, rem); + p->pos += rem; + *size = rem; + return SZ_OK; +} + +static SRes LookToRead2_Seek(const ILookInStream *pp, Int64 *pos, ESzSeek origin) +{ + GET_LookToRead2 + p->pos = p->size = 0; + return ISeekInStream_Seek(p->realStream, pos, origin); +} + +void LookToRead2_CreateVTable(CLookToRead2 *p, int lookahead) +{ + p->vt.Look = lookahead ? + LookToRead2_Look_Lookahead : + LookToRead2_Look_Exact; + p->vt.Skip = LookToRead2_Skip; + p->vt.Read = LookToRead2_Read; + p->vt.Seek = LookToRead2_Seek; +} + + + +static SRes SecToLook_Read(const ISeqInStream *pp, void *buf, size_t *size) +{ + CSecToLook *p = CONTAINER_FROM_VTBL(pp, CSecToLook, vt); + return LookInStream_LookRead(p->realStream, buf, size); +} + +void SecToLook_CreateVTable(CSecToLook *p) +{ + p->vt.Read = SecToLook_Read; +} + +static SRes SecToRead_Read(const ISeqInStream *pp, void *buf, size_t *size) +{ + CSecToRead *p = CONTAINER_FROM_VTBL(pp, CSecToRead, vt); + return ILookInStream_Read(p->realStream, buf, size); +} + +void SecToRead_CreateVTable(CSecToRead *p) +{ + p->vt.Read = SecToRead_Read; +} diff --git a/deps/LZMA-SDK/C/7zTypes.h b/deps/LZMA-SDK/C/7zTypes.h index 763d3a9b6..f7d707183 100644 --- a/deps/LZMA-SDK/C/7zTypes.h +++ b/deps/LZMA-SDK/C/7zTypes.h @@ -1,529 +1,529 @@ -/* 7zTypes.h -- Basic types -2022-04-01 : Igor Pavlov : Public domain */ - -#ifndef __7Z_TYPES_H -#define __7Z_TYPES_H - -#ifdef _WIN32 -/* #include */ -#else -#include -#endif - -#include - -#ifndef EXTERN_C_BEGIN -#ifdef __cplusplus -#define EXTERN_C_BEGIN extern "C" { -#define EXTERN_C_END } -#else -#define EXTERN_C_BEGIN -#define EXTERN_C_END -#endif -#endif - -EXTERN_C_BEGIN - -#define SZ_OK 0 - -#define SZ_ERROR_DATA 1 -#define SZ_ERROR_MEM 2 -#define SZ_ERROR_CRC 3 -#define SZ_ERROR_UNSUPPORTED 4 -#define SZ_ERROR_PARAM 5 -#define SZ_ERROR_INPUT_EOF 6 -#define SZ_ERROR_OUTPUT_EOF 7 -#define SZ_ERROR_READ 8 -#define SZ_ERROR_WRITE 9 -#define SZ_ERROR_PROGRESS 10 -#define SZ_ERROR_FAIL 11 -#define SZ_ERROR_THREAD 12 - -#define SZ_ERROR_ARCHIVE 16 -#define SZ_ERROR_NO_ARCHIVE 17 - -typedef int SRes; - - -#ifdef _MSC_VER - #if _MSC_VER > 1200 - #define MY_ALIGN(n) __declspec(align(n)) - #else - #define MY_ALIGN(n) - #endif -#else - #define MY_ALIGN(n) __attribute__ ((aligned(n))) -#endif - - -#ifdef _WIN32 - -/* typedef DWORD WRes; */ -typedef unsigned WRes; -#define MY_SRes_HRESULT_FROM_WRes(x) HRESULT_FROM_WIN32(x) - -// #define MY_HRES_ERROR__INTERNAL_ERROR MY_SRes_HRESULT_FROM_WRes(ERROR_INTERNAL_ERROR) - -#else // _WIN32 - -// #define ENV_HAVE_LSTAT -typedef int WRes; - -// (FACILITY_ERRNO = 0x800) is 7zip's FACILITY constant to represent (errno) errors in HRESULT -#define MY__FACILITY_ERRNO 0x800 -#define MY__FACILITY_WIN32 7 -#define MY__FACILITY__WRes MY__FACILITY_ERRNO - -#define MY_HRESULT_FROM_errno_CONST_ERROR(x) ((HRESULT)( \ - ( (HRESULT)(x) & 0x0000FFFF) \ - | (MY__FACILITY__WRes << 16) \ - | (HRESULT)0x80000000 )) - -#define MY_SRes_HRESULT_FROM_WRes(x) \ - ((HRESULT)(x) <= 0 ? ((HRESULT)(x)) : MY_HRESULT_FROM_errno_CONST_ERROR(x)) - -// we call macro HRESULT_FROM_WIN32 for system errors (WRes) that are (errno) -#define HRESULT_FROM_WIN32(x) MY_SRes_HRESULT_FROM_WRes(x) - -/* -#define ERROR_FILE_NOT_FOUND 2L -#define ERROR_ACCESS_DENIED 5L -#define ERROR_NO_MORE_FILES 18L -#define ERROR_LOCK_VIOLATION 33L -#define ERROR_FILE_EXISTS 80L -#define ERROR_DISK_FULL 112L -#define ERROR_NEGATIVE_SEEK 131L -#define ERROR_ALREADY_EXISTS 183L -#define ERROR_DIRECTORY 267L -#define ERROR_TOO_MANY_POSTS 298L - -#define ERROR_INTERNAL_ERROR 1359L -#define ERROR_INVALID_REPARSE_DATA 4392L -#define ERROR_REPARSE_TAG_INVALID 4393L -#define ERROR_REPARSE_TAG_MISMATCH 4394L -*/ - -// we use errno equivalents for some WIN32 errors: - -#define ERROR_INVALID_PARAMETER EINVAL -#define ERROR_INVALID_FUNCTION EINVAL -#define ERROR_ALREADY_EXISTS EEXIST -#define ERROR_FILE_EXISTS EEXIST -#define ERROR_PATH_NOT_FOUND ENOENT -#define ERROR_FILE_NOT_FOUND ENOENT -#define ERROR_DISK_FULL ENOSPC -// #define ERROR_INVALID_HANDLE EBADF - -// we use FACILITY_WIN32 for errors that has no errno equivalent -// Too many posts were made to a semaphore. -#define ERROR_TOO_MANY_POSTS ((HRESULT)0x8007012AL) -#define ERROR_INVALID_REPARSE_DATA ((HRESULT)0x80071128L) -#define ERROR_REPARSE_TAG_INVALID ((HRESULT)0x80071129L) - -// if (MY__FACILITY__WRes != FACILITY_WIN32), -// we use FACILITY_WIN32 for COM errors: -#define E_OUTOFMEMORY ((HRESULT)0x8007000EL) -#define E_INVALIDARG ((HRESULT)0x80070057L) -#define MY__E_ERROR_NEGATIVE_SEEK ((HRESULT)0x80070083L) - -/* -// we can use FACILITY_ERRNO for some COM errors, that have errno equivalents: -#define E_OUTOFMEMORY MY_HRESULT_FROM_errno_CONST_ERROR(ENOMEM) -#define E_INVALIDARG MY_HRESULT_FROM_errno_CONST_ERROR(EINVAL) -#define MY__E_ERROR_NEGATIVE_SEEK MY_HRESULT_FROM_errno_CONST_ERROR(EINVAL) -*/ - -#define TEXT(quote) quote - -#define FILE_ATTRIBUTE_READONLY 0x0001 -#define FILE_ATTRIBUTE_HIDDEN 0x0002 -#define FILE_ATTRIBUTE_SYSTEM 0x0004 -#define FILE_ATTRIBUTE_DIRECTORY 0x0010 -#define FILE_ATTRIBUTE_ARCHIVE 0x0020 -#define FILE_ATTRIBUTE_DEVICE 0x0040 -#define FILE_ATTRIBUTE_NORMAL 0x0080 -#define FILE_ATTRIBUTE_TEMPORARY 0x0100 -#define FILE_ATTRIBUTE_SPARSE_FILE 0x0200 -#define FILE_ATTRIBUTE_REPARSE_POINT 0x0400 -#define FILE_ATTRIBUTE_COMPRESSED 0x0800 -#define FILE_ATTRIBUTE_OFFLINE 0x1000 -#define FILE_ATTRIBUTE_NOT_CONTENT_INDEXED 0x2000 -#define FILE_ATTRIBUTE_ENCRYPTED 0x4000 - -#define FILE_ATTRIBUTE_UNIX_EXTENSION 0x8000 /* trick for Unix */ - -#endif - - -#ifndef RINOK -#define RINOK(x) { int __result__ = (x); if (__result__ != 0) return __result__; } -#endif - -#ifndef RINOK_WRes -#define RINOK_WRes(x) { WRes __result__ = (x); if (__result__ != 0) return __result__; } -#endif - -typedef unsigned char Byte; -typedef short Int16; -typedef unsigned short UInt16; - -#ifdef _LZMA_UINT32_IS_ULONG -typedef long Int32; -typedef unsigned long UInt32; -#else -typedef int Int32; -typedef unsigned int UInt32; -#endif - - -#ifndef _WIN32 - -typedef int INT; -typedef Int32 INT32; -typedef unsigned int UINT; -typedef UInt32 UINT32; -typedef INT32 LONG; // LONG, ULONG and DWORD must be 32-bit for _WIN32 compatibility -typedef UINT32 ULONG; - -#undef DWORD -typedef UINT32 DWORD; - -#define VOID void - -#define HRESULT LONG - -typedef void *LPVOID; -// typedef void VOID; -// typedef ULONG_PTR DWORD_PTR, *PDWORD_PTR; -// gcc / clang on Unix : sizeof(long==sizeof(void*) in 32 or 64 bits) -typedef long INT_PTR; -typedef unsigned long UINT_PTR; -typedef long LONG_PTR; -typedef unsigned long DWORD_PTR; - -typedef size_t SIZE_T; - -#endif // _WIN32 - - -#define MY_HRES_ERROR__INTERNAL_ERROR ((HRESULT)0x8007054FL) - - -#ifdef _SZ_NO_INT_64 - -/* define _SZ_NO_INT_64, if your compiler doesn't support 64-bit integers. - NOTES: Some code will work incorrectly in that case! */ - -typedef long Int64; -typedef unsigned long UInt64; - -#else - -#if defined(_MSC_VER) || defined(__BORLANDC__) -typedef __int64 Int64; -typedef unsigned __int64 UInt64; -#define UINT64_CONST(n) n -#else -typedef long long int Int64; -typedef unsigned long long int UInt64; -#define UINT64_CONST(n) n ## ULL -#endif - -#endif - -#ifdef _LZMA_NO_SYSTEM_SIZE_T -typedef UInt32 SizeT; -#else -typedef size_t SizeT; -#endif - -typedef int BoolInt; -/* typedef BoolInt Bool; */ -#define True 1 -#define False 0 - - -#ifdef _WIN32 -#define MY_STD_CALL __stdcall -#else -#define MY_STD_CALL -#endif - -#ifdef _MSC_VER - -#if _MSC_VER >= 1300 -#define MY_NO_INLINE __declspec(noinline) -#else -#define MY_NO_INLINE -#endif - -#define MY_FORCE_INLINE __forceinline - -#define MY_CDECL __cdecl -#define MY_FAST_CALL __fastcall - -#else // _MSC_VER - -#if (defined(__GNUC__) && (__GNUC__ >= 4)) \ - || (defined(__clang__) && (__clang_major__ >= 4)) \ - || defined(__INTEL_COMPILER) \ - || defined(__xlC__) -#define MY_NO_INLINE __attribute__((noinline)) -// #define MY_FORCE_INLINE __attribute__((always_inline)) inline -#else -#define MY_NO_INLINE -#endif - -#define MY_FORCE_INLINE - - -#define MY_CDECL - -#if defined(_M_IX86) \ - || defined(__i386__) -// #define MY_FAST_CALL __attribute__((fastcall)) -// #define MY_FAST_CALL __attribute__((cdecl)) -#define MY_FAST_CALL -#elif defined(MY_CPU_AMD64) -// #define MY_FAST_CALL __attribute__((ms_abi)) -#define MY_FAST_CALL -#else -#define MY_FAST_CALL -#endif - -#endif // _MSC_VER - - -/* The following interfaces use first parameter as pointer to structure */ - -typedef struct IByteIn IByteIn; -struct IByteIn -{ - Byte (*Read)(const IByteIn *p); /* reads one byte, returns 0 in case of EOF or error */ -}; -#define IByteIn_Read(p) (p)->Read(p) - - -typedef struct IByteOut IByteOut; -struct IByteOut -{ - void (*Write)(const IByteOut *p, Byte b); -}; -#define IByteOut_Write(p, b) (p)->Write(p, b) - - -typedef struct ISeqInStream ISeqInStream; -struct ISeqInStream -{ - SRes (*Read)(const ISeqInStream *p, void *buf, size_t *size); - /* if (input(*size) != 0 && output(*size) == 0) means end_of_stream. - (output(*size) < input(*size)) is allowed */ -}; -#define ISeqInStream_Read(p, buf, size) (p)->Read(p, buf, size) - -/* it can return SZ_ERROR_INPUT_EOF */ -SRes SeqInStream_Read(const ISeqInStream *stream, void *buf, size_t size); -SRes SeqInStream_Read2(const ISeqInStream *stream, void *buf, size_t size, SRes errorType); -SRes SeqInStream_ReadByte(const ISeqInStream *stream, Byte *buf); - - -typedef struct ISeqOutStream ISeqOutStream; -struct ISeqOutStream -{ - size_t (*Write)(const ISeqOutStream *p, const void *buf, size_t size); - /* Returns: result - the number of actually written bytes. - (result < size) means error */ -}; -#define ISeqOutStream_Write(p, buf, size) (p)->Write(p, buf, size) - -typedef enum -{ - SZ_SEEK_SET = 0, - SZ_SEEK_CUR = 1, - SZ_SEEK_END = 2 -} ESzSeek; - - -typedef struct ISeekInStream ISeekInStream; -struct ISeekInStream -{ - SRes (*Read)(const ISeekInStream *p, void *buf, size_t *size); /* same as ISeqInStream::Read */ - SRes (*Seek)(const ISeekInStream *p, Int64 *pos, ESzSeek origin); -}; -#define ISeekInStream_Read(p, buf, size) (p)->Read(p, buf, size) -#define ISeekInStream_Seek(p, pos, origin) (p)->Seek(p, pos, origin) - - -typedef struct ILookInStream ILookInStream; -struct ILookInStream -{ - SRes (*Look)(const ILookInStream *p, const void **buf, size_t *size); - /* if (input(*size) != 0 && output(*size) == 0) means end_of_stream. - (output(*size) > input(*size)) is not allowed - (output(*size) < input(*size)) is allowed */ - SRes (*Skip)(const ILookInStream *p, size_t offset); - /* offset must be <= output(*size) of Look */ - - SRes (*Read)(const ILookInStream *p, void *buf, size_t *size); - /* reads directly (without buffer). It's same as ISeqInStream::Read */ - SRes (*Seek)(const ILookInStream *p, Int64 *pos, ESzSeek origin); -}; - -#define ILookInStream_Look(p, buf, size) (p)->Look(p, buf, size) -#define ILookInStream_Skip(p, offset) (p)->Skip(p, offset) -#define ILookInStream_Read(p, buf, size) (p)->Read(p, buf, size) -#define ILookInStream_Seek(p, pos, origin) (p)->Seek(p, pos, origin) - - -SRes LookInStream_LookRead(const ILookInStream *stream, void *buf, size_t *size); -SRes LookInStream_SeekTo(const ILookInStream *stream, UInt64 offset); - -/* reads via ILookInStream::Read */ -SRes LookInStream_Read2(const ILookInStream *stream, void *buf, size_t size, SRes errorType); -SRes LookInStream_Read(const ILookInStream *stream, void *buf, size_t size); - - - -typedef struct -{ - ILookInStream vt; - const ISeekInStream *realStream; - - size_t pos; - size_t size; /* it's data size */ - - /* the following variables must be set outside */ - Byte *buf; - size_t bufSize; -} CLookToRead2; - -void LookToRead2_CreateVTable(CLookToRead2 *p, int lookahead); - -#define LookToRead2_Init(p) { (p)->pos = (p)->size = 0; } - - -typedef struct -{ - ISeqInStream vt; - const ILookInStream *realStream; -} CSecToLook; - -void SecToLook_CreateVTable(CSecToLook *p); - - - -typedef struct -{ - ISeqInStream vt; - const ILookInStream *realStream; -} CSecToRead; - -void SecToRead_CreateVTable(CSecToRead *p); - - -typedef struct ICompressProgress ICompressProgress; - -struct ICompressProgress -{ - SRes (*Progress)(const ICompressProgress *p, UInt64 inSize, UInt64 outSize); - /* Returns: result. (result != SZ_OK) means break. - Value (UInt64)(Int64)-1 for size means unknown value. */ -}; -#define ICompressProgress_Progress(p, inSize, outSize) (p)->Progress(p, inSize, outSize) - - - -typedef struct ISzAlloc ISzAlloc; -typedef const ISzAlloc * ISzAllocPtr; - -struct ISzAlloc -{ - void *(*Alloc)(ISzAllocPtr p, size_t size); - void (*Free)(ISzAllocPtr p, void *address); /* address can be 0 */ -}; - -#define ISzAlloc_Alloc(p, size) (p)->Alloc(p, size) -#define ISzAlloc_Free(p, a) (p)->Free(p, a) - -/* deprecated */ -#define IAlloc_Alloc(p, size) ISzAlloc_Alloc(p, size) -#define IAlloc_Free(p, a) ISzAlloc_Free(p, a) - - - - - -#ifndef MY_offsetof - #ifdef offsetof - #define MY_offsetof(type, m) offsetof(type, m) - /* - #define MY_offsetof(type, m) FIELD_OFFSET(type, m) - */ - #else - #define MY_offsetof(type, m) ((size_t)&(((type *)0)->m)) - #endif -#endif - - - -#ifndef MY_container_of - -/* -#define MY_container_of(ptr, type, m) container_of(ptr, type, m) -#define MY_container_of(ptr, type, m) CONTAINING_RECORD(ptr, type, m) -#define MY_container_of(ptr, type, m) ((type *)((char *)(ptr) - offsetof(type, m))) -#define MY_container_of(ptr, type, m) (&((type *)0)->m == (ptr), ((type *)(((char *)(ptr)) - MY_offsetof(type, m)))) -*/ - -/* - GCC shows warning: "perhaps the 'offsetof' macro was used incorrectly" - GCC 3.4.4 : classes with constructor - GCC 4.8.1 : classes with non-public variable members" -*/ - -#define MY_container_of(ptr, type, m) ((type *)(void *)((char *)(void *)(1 ? (ptr) : &((type *)0)->m) - MY_offsetof(type, m))) - -#endif - -#define CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m) ((type *)(void *)(ptr)) - -/* -#define CONTAINER_FROM_VTBL(ptr, type, m) CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m) -*/ -#define CONTAINER_FROM_VTBL(ptr, type, m) MY_container_of(ptr, type, m) - -#define CONTAINER_FROM_VTBL_CLS(ptr, type, m) CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m) -/* -#define CONTAINER_FROM_VTBL_CLS(ptr, type, m) CONTAINER_FROM_VTBL(ptr, type, m) -*/ - - -#define MY_memset_0_ARRAY(a) memset((a), 0, sizeof(a)) - -#ifdef _WIN32 - -#define CHAR_PATH_SEPARATOR '\\' -#define WCHAR_PATH_SEPARATOR L'\\' -#define STRING_PATH_SEPARATOR "\\" -#define WSTRING_PATH_SEPARATOR L"\\" - -#else - -#define CHAR_PATH_SEPARATOR '/' -#define WCHAR_PATH_SEPARATOR L'/' -#define STRING_PATH_SEPARATOR "/" -#define WSTRING_PATH_SEPARATOR L"/" - -#endif - -#define k_PropVar_TimePrec_0 0 -#define k_PropVar_TimePrec_Unix 1 -#define k_PropVar_TimePrec_DOS 2 -#define k_PropVar_TimePrec_HighPrec 3 -#define k_PropVar_TimePrec_Base 16 -#define k_PropVar_TimePrec_100ns (k_PropVar_TimePrec_Base + 7) -#define k_PropVar_TimePrec_1ns (k_PropVar_TimePrec_Base + 9) - -EXTERN_C_END - -#endif +/* 7zTypes.h -- Basic types +2022-04-01 : Igor Pavlov : Public domain */ + +#ifndef __7Z_TYPES_H +#define __7Z_TYPES_H + +#ifdef _WIN32 +/* #include */ +#else +#include +#endif + +#include + +#ifndef EXTERN_C_BEGIN +#ifdef __cplusplus +#define EXTERN_C_BEGIN extern "C" { +#define EXTERN_C_END } +#else +#define EXTERN_C_BEGIN +#define EXTERN_C_END +#endif +#endif + +EXTERN_C_BEGIN + +#define SZ_OK 0 + +#define SZ_ERROR_DATA 1 +#define SZ_ERROR_MEM 2 +#define SZ_ERROR_CRC 3 +#define SZ_ERROR_UNSUPPORTED 4 +#define SZ_ERROR_PARAM 5 +#define SZ_ERROR_INPUT_EOF 6 +#define SZ_ERROR_OUTPUT_EOF 7 +#define SZ_ERROR_READ 8 +#define SZ_ERROR_WRITE 9 +#define SZ_ERROR_PROGRESS 10 +#define SZ_ERROR_FAIL 11 +#define SZ_ERROR_THREAD 12 + +#define SZ_ERROR_ARCHIVE 16 +#define SZ_ERROR_NO_ARCHIVE 17 + +typedef int SRes; + + +#ifdef _MSC_VER + #if _MSC_VER > 1200 + #define MY_ALIGN(n) __declspec(align(n)) + #else + #define MY_ALIGN(n) + #endif +#else + #define MY_ALIGN(n) __attribute__ ((aligned(n))) +#endif + + +#ifdef _WIN32 + +/* typedef DWORD WRes; */ +typedef unsigned WRes; +#define MY_SRes_HRESULT_FROM_WRes(x) HRESULT_FROM_WIN32(x) + +// #define MY_HRES_ERROR__INTERNAL_ERROR MY_SRes_HRESULT_FROM_WRes(ERROR_INTERNAL_ERROR) + +#else // _WIN32 + +// #define ENV_HAVE_LSTAT +typedef int WRes; + +// (FACILITY_ERRNO = 0x800) is 7zip's FACILITY constant to represent (errno) errors in HRESULT +#define MY__FACILITY_ERRNO 0x800 +#define MY__FACILITY_WIN32 7 +#define MY__FACILITY__WRes MY__FACILITY_ERRNO + +#define MY_HRESULT_FROM_errno_CONST_ERROR(x) ((HRESULT)( \ + ( (HRESULT)(x) & 0x0000FFFF) \ + | (MY__FACILITY__WRes << 16) \ + | (HRESULT)0x80000000 )) + +#define MY_SRes_HRESULT_FROM_WRes(x) \ + ((HRESULT)(x) <= 0 ? ((HRESULT)(x)) : MY_HRESULT_FROM_errno_CONST_ERROR(x)) + +// we call macro HRESULT_FROM_WIN32 for system errors (WRes) that are (errno) +#define HRESULT_FROM_WIN32(x) MY_SRes_HRESULT_FROM_WRes(x) + +/* +#define ERROR_FILE_NOT_FOUND 2L +#define ERROR_ACCESS_DENIED 5L +#define ERROR_NO_MORE_FILES 18L +#define ERROR_LOCK_VIOLATION 33L +#define ERROR_FILE_EXISTS 80L +#define ERROR_DISK_FULL 112L +#define ERROR_NEGATIVE_SEEK 131L +#define ERROR_ALREADY_EXISTS 183L +#define ERROR_DIRECTORY 267L +#define ERROR_TOO_MANY_POSTS 298L + +#define ERROR_INTERNAL_ERROR 1359L +#define ERROR_INVALID_REPARSE_DATA 4392L +#define ERROR_REPARSE_TAG_INVALID 4393L +#define ERROR_REPARSE_TAG_MISMATCH 4394L +*/ + +// we use errno equivalents for some WIN32 errors: + +#define ERROR_INVALID_PARAMETER EINVAL +#define ERROR_INVALID_FUNCTION EINVAL +#define ERROR_ALREADY_EXISTS EEXIST +#define ERROR_FILE_EXISTS EEXIST +#define ERROR_PATH_NOT_FOUND ENOENT +#define ERROR_FILE_NOT_FOUND ENOENT +#define ERROR_DISK_FULL ENOSPC +// #define ERROR_INVALID_HANDLE EBADF + +// we use FACILITY_WIN32 for errors that has no errno equivalent +// Too many posts were made to a semaphore. +#define ERROR_TOO_MANY_POSTS ((HRESULT)0x8007012AL) +#define ERROR_INVALID_REPARSE_DATA ((HRESULT)0x80071128L) +#define ERROR_REPARSE_TAG_INVALID ((HRESULT)0x80071129L) + +// if (MY__FACILITY__WRes != FACILITY_WIN32), +// we use FACILITY_WIN32 for COM errors: +#define E_OUTOFMEMORY ((HRESULT)0x8007000EL) +#define E_INVALIDARG ((HRESULT)0x80070057L) +#define MY__E_ERROR_NEGATIVE_SEEK ((HRESULT)0x80070083L) + +/* +// we can use FACILITY_ERRNO for some COM errors, that have errno equivalents: +#define E_OUTOFMEMORY MY_HRESULT_FROM_errno_CONST_ERROR(ENOMEM) +#define E_INVALIDARG MY_HRESULT_FROM_errno_CONST_ERROR(EINVAL) +#define MY__E_ERROR_NEGATIVE_SEEK MY_HRESULT_FROM_errno_CONST_ERROR(EINVAL) +*/ + +#define TEXT(quote) quote + +#define FILE_ATTRIBUTE_READONLY 0x0001 +#define FILE_ATTRIBUTE_HIDDEN 0x0002 +#define FILE_ATTRIBUTE_SYSTEM 0x0004 +#define FILE_ATTRIBUTE_DIRECTORY 0x0010 +#define FILE_ATTRIBUTE_ARCHIVE 0x0020 +#define FILE_ATTRIBUTE_DEVICE 0x0040 +#define FILE_ATTRIBUTE_NORMAL 0x0080 +#define FILE_ATTRIBUTE_TEMPORARY 0x0100 +#define FILE_ATTRIBUTE_SPARSE_FILE 0x0200 +#define FILE_ATTRIBUTE_REPARSE_POINT 0x0400 +#define FILE_ATTRIBUTE_COMPRESSED 0x0800 +#define FILE_ATTRIBUTE_OFFLINE 0x1000 +#define FILE_ATTRIBUTE_NOT_CONTENT_INDEXED 0x2000 +#define FILE_ATTRIBUTE_ENCRYPTED 0x4000 + +#define FILE_ATTRIBUTE_UNIX_EXTENSION 0x8000 /* trick for Unix */ + +#endif + + +#ifndef RINOK +#define RINOK(x) { int __result__ = (x); if (__result__ != 0) return __result__; } +#endif + +#ifndef RINOK_WRes +#define RINOK_WRes(x) { WRes __result__ = (x); if (__result__ != 0) return __result__; } +#endif + +typedef unsigned char Byte; +typedef short Int16; +typedef unsigned short UInt16; + +#ifdef _LZMA_UINT32_IS_ULONG +typedef long Int32; +typedef unsigned long UInt32; +#else +typedef int Int32; +typedef unsigned int UInt32; +#endif + + +#ifndef _WIN32 + +typedef int INT; +typedef Int32 INT32; +typedef unsigned int UINT; +typedef UInt32 UINT32; +typedef INT32 LONG; // LONG, ULONG and DWORD must be 32-bit for _WIN32 compatibility +typedef UINT32 ULONG; + +#undef DWORD +typedef UINT32 DWORD; + +#define VOID void + +#define HRESULT LONG + +typedef void *LPVOID; +// typedef void VOID; +// typedef ULONG_PTR DWORD_PTR, *PDWORD_PTR; +// gcc / clang on Unix : sizeof(long==sizeof(void*) in 32 or 64 bits) +typedef long INT_PTR; +typedef unsigned long UINT_PTR; +typedef long LONG_PTR; +typedef unsigned long DWORD_PTR; + +typedef size_t SIZE_T; + +#endif // _WIN32 + + +#define MY_HRES_ERROR__INTERNAL_ERROR ((HRESULT)0x8007054FL) + + +#ifdef _SZ_NO_INT_64 + +/* define _SZ_NO_INT_64, if your compiler doesn't support 64-bit integers. + NOTES: Some code will work incorrectly in that case! */ + +typedef long Int64; +typedef unsigned long UInt64; + +#else + +#if defined(_MSC_VER) || defined(__BORLANDC__) +typedef __int64 Int64; +typedef unsigned __int64 UInt64; +#define UINT64_CONST(n) n +#else +typedef long long int Int64; +typedef unsigned long long int UInt64; +#define UINT64_CONST(n) n ## ULL +#endif + +#endif + +#ifdef _LZMA_NO_SYSTEM_SIZE_T +typedef UInt32 SizeT; +#else +typedef size_t SizeT; +#endif + +typedef int BoolInt; +/* typedef BoolInt Bool; */ +#define True 1 +#define False 0 + + +#ifdef _WIN32 +#define MY_STD_CALL __stdcall +#else +#define MY_STD_CALL +#endif + +#ifdef _MSC_VER + +#if _MSC_VER >= 1300 +#define MY_NO_INLINE __declspec(noinline) +#else +#define MY_NO_INLINE +#endif + +#define MY_FORCE_INLINE __forceinline + +#define MY_CDECL __cdecl +#define MY_FAST_CALL __fastcall + +#else // _MSC_VER + +#if (defined(__GNUC__) && (__GNUC__ >= 4)) \ + || (defined(__clang__) && (__clang_major__ >= 4)) \ + || defined(__INTEL_COMPILER) \ + || defined(__xlC__) +#define MY_NO_INLINE __attribute__((noinline)) +// #define MY_FORCE_INLINE __attribute__((always_inline)) inline +#else +#define MY_NO_INLINE +#endif + +#define MY_FORCE_INLINE + + +#define MY_CDECL + +#if defined(_M_IX86) \ + || defined(__i386__) +// #define MY_FAST_CALL __attribute__((fastcall)) +// #define MY_FAST_CALL __attribute__((cdecl)) +#define MY_FAST_CALL +#elif defined(MY_CPU_AMD64) +// #define MY_FAST_CALL __attribute__((ms_abi)) +#define MY_FAST_CALL +#else +#define MY_FAST_CALL +#endif + +#endif // _MSC_VER + + +/* The following interfaces use first parameter as pointer to structure */ + +typedef struct IByteIn IByteIn; +struct IByteIn +{ + Byte (*Read)(const IByteIn *p); /* reads one byte, returns 0 in case of EOF or error */ +}; +#define IByteIn_Read(p) (p)->Read(p) + + +typedef struct IByteOut IByteOut; +struct IByteOut +{ + void (*Write)(const IByteOut *p, Byte b); +}; +#define IByteOut_Write(p, b) (p)->Write(p, b) + + +typedef struct ISeqInStream ISeqInStream; +struct ISeqInStream +{ + SRes (*Read)(const ISeqInStream *p, void *buf, size_t *size); + /* if (input(*size) != 0 && output(*size) == 0) means end_of_stream. + (output(*size) < input(*size)) is allowed */ +}; +#define ISeqInStream_Read(p, buf, size) (p)->Read(p, buf, size) + +/* it can return SZ_ERROR_INPUT_EOF */ +SRes SeqInStream_Read(const ISeqInStream *stream, void *buf, size_t size); +SRes SeqInStream_Read2(const ISeqInStream *stream, void *buf, size_t size, SRes errorType); +SRes SeqInStream_ReadByte(const ISeqInStream *stream, Byte *buf); + + +typedef struct ISeqOutStream ISeqOutStream; +struct ISeqOutStream +{ + size_t (*Write)(const ISeqOutStream *p, const void *buf, size_t size); + /* Returns: result - the number of actually written bytes. + (result < size) means error */ +}; +#define ISeqOutStream_Write(p, buf, size) (p)->Write(p, buf, size) + +typedef enum +{ + SZ_SEEK_SET = 0, + SZ_SEEK_CUR = 1, + SZ_SEEK_END = 2 +} ESzSeek; + + +typedef struct ISeekInStream ISeekInStream; +struct ISeekInStream +{ + SRes (*Read)(const ISeekInStream *p, void *buf, size_t *size); /* same as ISeqInStream::Read */ + SRes (*Seek)(const ISeekInStream *p, Int64 *pos, ESzSeek origin); +}; +#define ISeekInStream_Read(p, buf, size) (p)->Read(p, buf, size) +#define ISeekInStream_Seek(p, pos, origin) (p)->Seek(p, pos, origin) + + +typedef struct ILookInStream ILookInStream; +struct ILookInStream +{ + SRes (*Look)(const ILookInStream *p, const void **buf, size_t *size); + /* if (input(*size) != 0 && output(*size) == 0) means end_of_stream. + (output(*size) > input(*size)) is not allowed + (output(*size) < input(*size)) is allowed */ + SRes (*Skip)(const ILookInStream *p, size_t offset); + /* offset must be <= output(*size) of Look */ + + SRes (*Read)(const ILookInStream *p, void *buf, size_t *size); + /* reads directly (without buffer). It's same as ISeqInStream::Read */ + SRes (*Seek)(const ILookInStream *p, Int64 *pos, ESzSeek origin); +}; + +#define ILookInStream_Look(p, buf, size) (p)->Look(p, buf, size) +#define ILookInStream_Skip(p, offset) (p)->Skip(p, offset) +#define ILookInStream_Read(p, buf, size) (p)->Read(p, buf, size) +#define ILookInStream_Seek(p, pos, origin) (p)->Seek(p, pos, origin) + + +SRes LookInStream_LookRead(const ILookInStream *stream, void *buf, size_t *size); +SRes LookInStream_SeekTo(const ILookInStream *stream, UInt64 offset); + +/* reads via ILookInStream::Read */ +SRes LookInStream_Read2(const ILookInStream *stream, void *buf, size_t size, SRes errorType); +SRes LookInStream_Read(const ILookInStream *stream, void *buf, size_t size); + + + +typedef struct +{ + ILookInStream vt; + const ISeekInStream *realStream; + + size_t pos; + size_t size; /* it's data size */ + + /* the following variables must be set outside */ + Byte *buf; + size_t bufSize; +} CLookToRead2; + +void LookToRead2_CreateVTable(CLookToRead2 *p, int lookahead); + +#define LookToRead2_Init(p) { (p)->pos = (p)->size = 0; } + + +typedef struct +{ + ISeqInStream vt; + const ILookInStream *realStream; +} CSecToLook; + +void SecToLook_CreateVTable(CSecToLook *p); + + + +typedef struct +{ + ISeqInStream vt; + const ILookInStream *realStream; +} CSecToRead; + +void SecToRead_CreateVTable(CSecToRead *p); + + +typedef struct ICompressProgress ICompressProgress; + +struct ICompressProgress +{ + SRes (*Progress)(const ICompressProgress *p, UInt64 inSize, UInt64 outSize); + /* Returns: result. (result != SZ_OK) means break. + Value (UInt64)(Int64)-1 for size means unknown value. */ +}; +#define ICompressProgress_Progress(p, inSize, outSize) (p)->Progress(p, inSize, outSize) + + + +typedef struct ISzAlloc ISzAlloc; +typedef const ISzAlloc * ISzAllocPtr; + +struct ISzAlloc +{ + void *(*Alloc)(ISzAllocPtr p, size_t size); + void (*Free)(ISzAllocPtr p, void *address); /* address can be 0 */ +}; + +#define ISzAlloc_Alloc(p, size) (p)->Alloc(p, size) +#define ISzAlloc_Free(p, a) (p)->Free(p, a) + +/* deprecated */ +#define IAlloc_Alloc(p, size) ISzAlloc_Alloc(p, size) +#define IAlloc_Free(p, a) ISzAlloc_Free(p, a) + + + + + +#ifndef MY_offsetof + #ifdef offsetof + #define MY_offsetof(type, m) offsetof(type, m) + /* + #define MY_offsetof(type, m) FIELD_OFFSET(type, m) + */ + #else + #define MY_offsetof(type, m) ((size_t)&(((type *)0)->m)) + #endif +#endif + + + +#ifndef MY_container_of + +/* +#define MY_container_of(ptr, type, m) container_of(ptr, type, m) +#define MY_container_of(ptr, type, m) CONTAINING_RECORD(ptr, type, m) +#define MY_container_of(ptr, type, m) ((type *)((char *)(ptr) - offsetof(type, m))) +#define MY_container_of(ptr, type, m) (&((type *)0)->m == (ptr), ((type *)(((char *)(ptr)) - MY_offsetof(type, m)))) +*/ + +/* + GCC shows warning: "perhaps the 'offsetof' macro was used incorrectly" + GCC 3.4.4 : classes with constructor + GCC 4.8.1 : classes with non-public variable members" +*/ + +#define MY_container_of(ptr, type, m) ((type *)(void *)((char *)(void *)(1 ? (ptr) : &((type *)0)->m) - MY_offsetof(type, m))) + +#endif + +#define CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m) ((type *)(void *)(ptr)) + +/* +#define CONTAINER_FROM_VTBL(ptr, type, m) CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m) +*/ +#define CONTAINER_FROM_VTBL(ptr, type, m) MY_container_of(ptr, type, m) + +#define CONTAINER_FROM_VTBL_CLS(ptr, type, m) CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m) +/* +#define CONTAINER_FROM_VTBL_CLS(ptr, type, m) CONTAINER_FROM_VTBL(ptr, type, m) +*/ + + +#define MY_memset_0_ARRAY(a) memset((a), 0, sizeof(a)) + +#ifdef _WIN32 + +#define CHAR_PATH_SEPARATOR '\\' +#define WCHAR_PATH_SEPARATOR L'\\' +#define STRING_PATH_SEPARATOR "\\" +#define WSTRING_PATH_SEPARATOR L"\\" + +#else + +#define CHAR_PATH_SEPARATOR '/' +#define WCHAR_PATH_SEPARATOR L'/' +#define STRING_PATH_SEPARATOR "/" +#define WSTRING_PATH_SEPARATOR L"/" + +#endif + +#define k_PropVar_TimePrec_0 0 +#define k_PropVar_TimePrec_Unix 1 +#define k_PropVar_TimePrec_DOS 2 +#define k_PropVar_TimePrec_HighPrec 3 +#define k_PropVar_TimePrec_Base 16 +#define k_PropVar_TimePrec_100ns (k_PropVar_TimePrec_Base + 7) +#define k_PropVar_TimePrec_1ns (k_PropVar_TimePrec_Base + 9) + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/7zVersion.h b/deps/LZMA-SDK/C/7zVersion.h index fa9e6fc53..49ea81dd3 100644 --- a/deps/LZMA-SDK/C/7zVersion.h +++ b/deps/LZMA-SDK/C/7zVersion.h @@ -1,27 +1,27 @@ -#define MY_VER_MAJOR 22 -#define MY_VER_MINOR 01 -#define MY_VER_BUILD 0 -#define MY_VERSION_NUMBERS "22.01" -#define MY_VERSION MY_VERSION_NUMBERS - -#ifdef MY_CPU_NAME - #define MY_VERSION_CPU MY_VERSION " (" MY_CPU_NAME ")" -#else - #define MY_VERSION_CPU MY_VERSION -#endif - -#define MY_DATE "2022-07-15" -#undef MY_COPYRIGHT -#undef MY_VERSION_COPYRIGHT_DATE -#define MY_AUTHOR_NAME "Igor Pavlov" -#define MY_COPYRIGHT_PD "Igor Pavlov : Public domain" -#define MY_COPYRIGHT_CR "Copyright (c) 1999-2022 Igor Pavlov" - -#ifdef USE_COPYRIGHT_CR - #define MY_COPYRIGHT MY_COPYRIGHT_CR -#else - #define MY_COPYRIGHT MY_COPYRIGHT_PD -#endif - -#define MY_COPYRIGHT_DATE MY_COPYRIGHT " : " MY_DATE -#define MY_VERSION_COPYRIGHT_DATE MY_VERSION_CPU " : " MY_COPYRIGHT " : " MY_DATE +#define MY_VER_MAJOR 22 +#define MY_VER_MINOR 01 +#define MY_VER_BUILD 0 +#define MY_VERSION_NUMBERS "22.01" +#define MY_VERSION MY_VERSION_NUMBERS + +#ifdef MY_CPU_NAME + #define MY_VERSION_CPU MY_VERSION " (" MY_CPU_NAME ")" +#else + #define MY_VERSION_CPU MY_VERSION +#endif + +#define MY_DATE "2022-07-15" +#undef MY_COPYRIGHT +#undef MY_VERSION_COPYRIGHT_DATE +#define MY_AUTHOR_NAME "Igor Pavlov" +#define MY_COPYRIGHT_PD "Igor Pavlov : Public domain" +#define MY_COPYRIGHT_CR "Copyright (c) 1999-2022 Igor Pavlov" + +#ifdef USE_COPYRIGHT_CR + #define MY_COPYRIGHT MY_COPYRIGHT_CR +#else + #define MY_COPYRIGHT MY_COPYRIGHT_PD +#endif + +#define MY_COPYRIGHT_DATE MY_COPYRIGHT " : " MY_DATE +#define MY_VERSION_COPYRIGHT_DATE MY_VERSION_CPU " : " MY_COPYRIGHT " : " MY_DATE diff --git a/deps/LZMA-SDK/C/7zVersion.rc b/deps/LZMA-SDK/C/7zVersion.rc index 6ed26de74..e520995dd 100644 --- a/deps/LZMA-SDK/C/7zVersion.rc +++ b/deps/LZMA-SDK/C/7zVersion.rc @@ -1,55 +1,55 @@ -#define MY_VS_FFI_FILEFLAGSMASK 0x0000003FL -#define MY_VOS_NT_WINDOWS32 0x00040004L -#define MY_VOS_CE_WINDOWS32 0x00050004L - -#define MY_VFT_APP 0x00000001L -#define MY_VFT_DLL 0x00000002L - -// #include - -#ifndef MY_VERSION -#include "7zVersion.h" -#endif - -#define MY_VER MY_VER_MAJOR,MY_VER_MINOR,MY_VER_BUILD,0 - -#ifdef DEBUG -#define DBG_FL VS_FF_DEBUG -#else -#define DBG_FL 0 -#endif - -#define MY_VERSION_INFO(fileType, descr, intName, origName) \ -LANGUAGE 9, 1 \ -1 VERSIONINFO \ - FILEVERSION MY_VER \ - PRODUCTVERSION MY_VER \ - FILEFLAGSMASK MY_VS_FFI_FILEFLAGSMASK \ - FILEFLAGS DBG_FL \ - FILEOS MY_VOS_NT_WINDOWS32 \ - FILETYPE fileType \ - FILESUBTYPE 0x0L \ -BEGIN \ - BLOCK "StringFileInfo" \ - BEGIN \ - BLOCK "040904b0" \ - BEGIN \ - VALUE "CompanyName", "Igor Pavlov" \ - VALUE "FileDescription", descr \ - VALUE "FileVersion", MY_VERSION \ - VALUE "InternalName", intName \ - VALUE "LegalCopyright", MY_COPYRIGHT \ - VALUE "OriginalFilename", origName \ - VALUE "ProductName", "7-Zip" \ - VALUE "ProductVersion", MY_VERSION \ - END \ - END \ - BLOCK "VarFileInfo" \ - BEGIN \ - VALUE "Translation", 0x409, 1200 \ - END \ -END - -#define MY_VERSION_INFO_APP(descr, intName) MY_VERSION_INFO(MY_VFT_APP, descr, intName, intName ".exe") - -#define MY_VERSION_INFO_DLL(descr, intName) MY_VERSION_INFO(MY_VFT_DLL, descr, intName, intName ".dll") +#define MY_VS_FFI_FILEFLAGSMASK 0x0000003FL +#define MY_VOS_NT_WINDOWS32 0x00040004L +#define MY_VOS_CE_WINDOWS32 0x00050004L + +#define MY_VFT_APP 0x00000001L +#define MY_VFT_DLL 0x00000002L + +// #include + +#ifndef MY_VERSION +#include "7zVersion.h" +#endif + +#define MY_VER MY_VER_MAJOR,MY_VER_MINOR,MY_VER_BUILD,0 + +#ifdef DEBUG +#define DBG_FL VS_FF_DEBUG +#else +#define DBG_FL 0 +#endif + +#define MY_VERSION_INFO(fileType, descr, intName, origName) \ +LANGUAGE 9, 1 \ +1 VERSIONINFO \ + FILEVERSION MY_VER \ + PRODUCTVERSION MY_VER \ + FILEFLAGSMASK MY_VS_FFI_FILEFLAGSMASK \ + FILEFLAGS DBG_FL \ + FILEOS MY_VOS_NT_WINDOWS32 \ + FILETYPE fileType \ + FILESUBTYPE 0x0L \ +BEGIN \ + BLOCK "StringFileInfo" \ + BEGIN \ + BLOCK "040904b0" \ + BEGIN \ + VALUE "CompanyName", "Igor Pavlov" \ + VALUE "FileDescription", descr \ + VALUE "FileVersion", MY_VERSION \ + VALUE "InternalName", intName \ + VALUE "LegalCopyright", MY_COPYRIGHT \ + VALUE "OriginalFilename", origName \ + VALUE "ProductName", "7-Zip" \ + VALUE "ProductVersion", MY_VERSION \ + END \ + END \ + BLOCK "VarFileInfo" \ + BEGIN \ + VALUE "Translation", 0x409, 1200 \ + END \ +END + +#define MY_VERSION_INFO_APP(descr, intName) MY_VERSION_INFO(MY_VFT_APP, descr, intName, intName ".exe") + +#define MY_VERSION_INFO_DLL(descr, intName) MY_VERSION_INFO(MY_VFT_DLL, descr, intName, intName ".dll") diff --git a/deps/LZMA-SDK/C/7zip_gcc_c.mak b/deps/LZMA-SDK/C/7zip_gcc_c.mak index d41810478..24505f377 100644 --- a/deps/LZMA-SDK/C/7zip_gcc_c.mak +++ b/deps/LZMA-SDK/C/7zip_gcc_c.mak @@ -1,343 +1,343 @@ - -MY_ARCH_2 = $(MY_ARCH) - -MY_ASM = jwasm -MY_ASM = asmc - -PROGPATH = $(O)/$(PROG) -PROGPATH_STATIC = $(O)/$(PROG)s - - -# for object file -CFLAGS_BASE_LIST = -c -# for ASM file -# CFLAGS_BASE_LIST = -S -CFLAGS_BASE = $(MY_ARCH_2) -O2 $(CFLAGS_BASE_LIST) -Wall -Werror -Wextra $(CFLAGS_WARN) \ - -DNDEBUG -D_REENTRANT -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE - - -ifdef SystemDrive -IS_MINGW = 1 -else -ifdef SYSTEMDRIVE -# ifdef OS -IS_MINGW = 1 -endif -endif - -ifdef IS_MINGW -LDFLAGS_STATIC_2 = -static -else -ifndef DEF_FILE -ifndef IS_NOT_STANDALONE -ifndef MY_DYNAMIC_LINK -ifneq ($(CC), clang) -LDFLAGS_STATIC_2 = -# -static -# -static-libstdc++ -static-libgcc -endif -endif -endif -endif -endif - -LDFLAGS_STATIC = -DNDEBUG $(LDFLAGS_STATIC_2) - -ifdef DEF_FILE - - -ifdef IS_MINGW -SHARED_EXT=.dll -LDFLAGS = -shared -DEF $(DEF_FILE) $(LDFLAGS_STATIC) -else -SHARED_EXT=.so -LDFLAGS = -shared -fPIC $(LDFLAGS_STATIC) -CC_SHARED=-fPIC -endif - - -else - -LDFLAGS = $(LDFLAGS_STATIC) -# -s is not required for clang, do we need it for GGC ??? -# -s - -#-static -static-libgcc -static-libstdc++ - -ifdef IS_MINGW -SHARED_EXT=.exe -else -SHARED_EXT= -endif - -endif - - -PROGPATH = $(O)/$(PROG)$(SHARED_EXT) -PROGPATH_STATIC = $(O)/$(PROG)s$(SHARED_EXT) - -ifndef O -O=_o -endif - -ifdef IS_MINGW - -ifdef MSYSTEM -RM = rm -f -MY_MKDIR=mkdir -p -DEL_OBJ_EXE = -$(RM) $(PROGPATH) $(PROGPATH_STATIC) $(OBJS) -else -RM = del -MY_MKDIR=mkdir -DEL_OBJ_EXE = -$(RM) $(O)\*.o $(O)\$(PROG).exe $(O)\$(PROG).dll -endif - - -LIB2 = -lOle32 -loleaut32 -luuid -ladvapi32 -lUser32 - -CXXFLAGS_EXTRA = -DUNICODE -D_UNICODE -# -Wno-delete-non-virtual-dtor - - -else - -RM = rm -f -MY_MKDIR=mkdir -p -# CFLAGS_BASE := $(CFLAGS_BASE) -D_7ZIP_ST -# CXXFLAGS_EXTRA = -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE - -# LOCAL_LIBS=-lpthread -# LOCAL_LIBS_DLL=$(LOCAL_LIBS) -ldl -LIB2 = -lpthread -ldl - -DEL_OBJ_EXE = -$(RM) $(PROGPATH) $(PROGPATH_STATIC) $(OBJS) - -endif - - - -CFLAGS = $(LOCAL_FLAGS) $(CFLAGS_BASE2) $(CFLAGS_BASE) $(CC_SHARED) -o $@ - - -ifdef IS_X64 -AFLAGS_ABI = -elf64 -DABI_LINUX -else -AFLAGS_ABI = -elf -DABI_LINUX -DABI_CDECL -# -DABI_CDECL -# -DABI_LINUX -# -DABI_CDECL -endif -AFLAGS = $(AFLAGS_ABI) -Fo$(O)/ - - -CXX_WARN_FLAGS = -#-Wno-invalid-offsetof -#-Wno-reorder - -CXXFLAGS = $(LOCAL_FLAGS) $(CXXFLAGS_BASE2) $(CFLAGS_BASE) $(CXXFLAGS_EXTRA) $(CC_SHARED) -o $@ $(CXX_WARN_FLAGS) - -STATIC_TARGET= -ifdef COMPL_STATIC -STATIC_TARGET=$(PROGPATH_STATIC) -endif - - -all: $(O) $(PROGPATH) $(STATIC_TARGET) - -$(O): - $(MY_MKDIR) $(O) - -LFLAGS_ALL = -s $(MY_ARCH_2) $(LDFLAGS) $(LD_arch) $(OBJS) $(MY_LIBS) $(LIB2) -$(PROGPATH): $(OBJS) - $(CXX) -o $(PROGPATH) $(LFLAGS_ALL) - -$(PROGPATH_STATIC): $(OBJS) - $(CXX) -static -o $(PROGPATH_STATIC) $(LFLAGS_ALL) - - -ifndef NO_DEFAULT_RES -$O/resource.o: resource.rc - windres.exe $(RFLAGS) resource.rc $O/resource.o -endif - - - -$O/7zAlloc.o: ../../../C/7zAlloc.c - $(CC) $(CFLAGS) $< -$O/7zArcIn.o: ../../../C/7zArcIn.c - $(CC) $(CFLAGS) $< -$O/7zBuf.o: ../../../C/7zBuf.c - $(CC) $(CFLAGS) $< -$O/7zBuf2.o: ../../../C/7zBuf2.c - $(CC) $(CFLAGS) $< -$O/7zCrc.o: ../../../C/7zCrc.c - $(CC) $(CFLAGS) $< -$O/7zDec.o: ../../../C/7zDec.c - $(CC) $(CFLAGS) $< -$O/7zFile.o: ../../../C/7zFile.c - $(CC) $(CFLAGS) $< -$O/7zStream.o: ../../../C/7zStream.c - $(CC) $(CFLAGS) $< -$O/Aes.o: ../../../C/Aes.c - $(CC) $(CFLAGS) $< -$O/Alloc.o: ../../../C/Alloc.c - $(CC) $(CFLAGS) $< -$O/Bcj2.o: ../../../C/Bcj2.c - $(CC) $(CFLAGS) $< -$O/Bcj2Enc.o: ../../../C/Bcj2Enc.c - $(CC) $(CFLAGS) $< -$O/Blake2s.o: ../../../C/Blake2s.c - $(CC) $(CFLAGS) $< -$O/Bra.o: ../../../C/Bra.c - $(CC) $(CFLAGS) $< -$O/Bra86.o: ../../../C/Bra86.c - $(CC) $(CFLAGS) $< -$O/BraIA64.o: ../../../C/BraIA64.c - $(CC) $(CFLAGS) $< -$O/BwtSort.o: ../../../C/BwtSort.c - $(CC) $(CFLAGS) $< - -$O/CpuArch.o: ../../../C/CpuArch.c - $(CC) $(CFLAGS) $< -$O/Delta.o: ../../../C/Delta.c - $(CC) $(CFLAGS) $< -$O/DllSecur.o: ../../../C/DllSecur.c - $(CC) $(CFLAGS) $< -$O/HuffEnc.o: ../../../C/HuffEnc.c - $(CC) $(CFLAGS) $< -$O/LzFind.o: ../../../C/LzFind.c - $(CC) $(CFLAGS) $< - -# ifdef MT_FILES -$O/LzFindMt.o: ../../../C/LzFindMt.c - $(CC) $(CFLAGS) $< -$O/LzFindOpt.o: ../../../C/LzFindOpt.c - $(CC) $(CFLAGS) $< - -$O/Threads.o: ../../../C/Threads.c - $(CC) $(CFLAGS) $< -# endif - -$O/LzmaEnc.o: ../../../C/LzmaEnc.c - $(CC) $(CFLAGS) $< -$O/Lzma86Dec.o: ../../../C/Lzma86Dec.c - $(CC) $(CFLAGS) $< -$O/Lzma86Enc.o: ../../../C/Lzma86Enc.c - $(CC) $(CFLAGS) $< -$O/Lzma2Dec.o: ../../../C/Lzma2Dec.c - $(CC) $(CFLAGS) $< -$O/Lzma2DecMt.o: ../../../C/Lzma2DecMt.c - $(CC) $(CFLAGS) $< -$O/Lzma2Enc.o: ../../../C/Lzma2Enc.c - $(CC) $(CFLAGS) $< -$O/LzmaLib.o: ../../../C/LzmaLib.c - $(CC) $(CFLAGS) $< -$O/MtCoder.o: ../../../C/MtCoder.c - $(CC) $(CFLAGS) $< -$O/MtDec.o: ../../../C/MtDec.c - $(CC) $(CFLAGS) $< -$O/Ppmd7.o: ../../../C/Ppmd7.c - $(CC) $(CFLAGS) $< -$O/Ppmd7aDec.o: ../../../C/Ppmd7aDec.c - $(CC) $(CFLAGS) $< -$O/Ppmd7Dec.o: ../../../C/Ppmd7Dec.c - $(CC) $(CFLAGS) $< -$O/Ppmd7Enc.o: ../../../C/Ppmd7Enc.c - $(CC) $(CFLAGS) $< -$O/Ppmd8.o: ../../../C/Ppmd8.c - $(CC) $(CFLAGS) $< -$O/Ppmd8Dec.o: ../../../C/Ppmd8Dec.c - $(CC) $(CFLAGS) $< -$O/Ppmd8Enc.o: ../../../C/Ppmd8Enc.c - $(CC) $(CFLAGS) $< -$O/Sha1.o: ../../../C/Sha1.c - $(CC) $(CFLAGS) $< -$O/Sha256.o: ../../../C/Sha256.c - $(CC) $(CFLAGS) $< -$O/Sort.o: ../../../C/Sort.c - $(CC) $(CFLAGS) $< -$O/Xz.o: ../../../C/Xz.c - $(CC) $(CFLAGS) $< -$O/XzCrc64.o: ../../../C/XzCrc64.c - $(CC) $(CFLAGS) $< - - -ifdef USE_ASM -ifdef IS_X64 -USE_X86_ASM=1 -else -ifdef IS_X86 -USE_X86_ASM=1 -endif -endif -endif - -ifdef USE_X86_ASM -$O/7zCrcOpt.o: ../../../Asm/x86/7zCrcOpt.asm - $(MY_ASM) $(AFLAGS) $< -$O/XzCrc64Opt.o: ../../../Asm/x86/XzCrc64Opt.asm - $(MY_ASM) $(AFLAGS) $< -$O/AesOpt.o: ../../../Asm/x86/AesOpt.asm - $(MY_ASM) $(AFLAGS) $< -$O/Sha1Opt.o: ../../../Asm/x86/Sha1Opt.asm - $(MY_ASM) $(AFLAGS) $< -$O/Sha256Opt.o: ../../../Asm/x86/Sha256Opt.asm - $(MY_ASM) $(AFLAGS) $< -else -$O/7zCrcOpt.o: ../../7zCrcOpt.c - $(CC) $(CFLAGS) $< -$O/XzCrc64Opt.o: ../../XzCrc64Opt.c - $(CC) $(CFLAGS) $< -$O/Sha1Opt.o: ../../Sha1Opt.c - $(CC) $(CFLAGS) $< -$O/Sha256Opt.o: ../../Sha256Opt.c - $(CC) $(CFLAGS) $< -$O/AesOpt.o: ../../AesOpt.c - $(CC) $(CFLAGS) $< -endif - - -ifdef USE_LZMA_DEC_ASM - -ifdef IS_X64 -$O/LzmaDecOpt.o: ../../../Asm/x86/LzmaDecOpt.asm - $(MY_ASM) $(AFLAGS) $< -endif - -ifdef IS_ARM64 -$O/LzmaDecOpt.o: ../../../Asm/arm64/LzmaDecOpt.S ../../../Asm/arm64/7zAsm.S - $(CC) $(CFLAGS) $< -endif - -$O/LzmaDec.o: ../../LzmaDec.c - $(CC) $(CFLAGS) -D_LZMA_DEC_OPT $< - -else - -$O/LzmaDec.o: ../../LzmaDec.c - $(CC) $(CFLAGS) $< - -endif - - - -$O/XzDec.o: ../../../C/XzDec.c - $(CC) $(CFLAGS) $< -$O/XzEnc.o: ../../../C/XzEnc.c - $(CC) $(CFLAGS) $< -$O/XzIn.o: ../../../C/XzIn.c - $(CC) $(CFLAGS) $< - - -$O/7zMain.o: ../../../C/Util/7z/7zMain.c - $(CC) $(CFLAGS) $< -$O/LzmaUtil.o: ../../../C/Util/Lzma/LzmaUtil.c - $(CC) $(CFLAGS) $< -$O/7zipInstall.o: ../../../C/Util/7zipInstall/7zipInstall.c - $(CC) $(CFLAGS) $< -$O/7zipUninstall.o: ../../../C/Util/7zipUninstall/7zipUninstall.c - $(CC) $(CFLAGS) $< - - -clean: - -$(DEL_OBJ_EXE) + +MY_ARCH_2 = $(MY_ARCH) + +MY_ASM = jwasm +MY_ASM = asmc + +PROGPATH = $(O)/$(PROG) +PROGPATH_STATIC = $(O)/$(PROG)s + + +# for object file +CFLAGS_BASE_LIST = -c +# for ASM file +# CFLAGS_BASE_LIST = -S +CFLAGS_BASE = $(MY_ARCH_2) -O2 $(CFLAGS_BASE_LIST) -Wall -Werror -Wextra $(CFLAGS_WARN) \ + -DNDEBUG -D_REENTRANT -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE + + +ifdef SystemDrive +IS_MINGW = 1 +else +ifdef SYSTEMDRIVE +# ifdef OS +IS_MINGW = 1 +endif +endif + +ifdef IS_MINGW +LDFLAGS_STATIC_2 = -static +else +ifndef DEF_FILE +ifndef IS_NOT_STANDALONE +ifndef MY_DYNAMIC_LINK +ifneq ($(CC), clang) +LDFLAGS_STATIC_2 = +# -static +# -static-libstdc++ -static-libgcc +endif +endif +endif +endif +endif + +LDFLAGS_STATIC = -DNDEBUG $(LDFLAGS_STATIC_2) + +ifdef DEF_FILE + + +ifdef IS_MINGW +SHARED_EXT=.dll +LDFLAGS = -shared -DEF $(DEF_FILE) $(LDFLAGS_STATIC) +else +SHARED_EXT=.so +LDFLAGS = -shared -fPIC $(LDFLAGS_STATIC) +CC_SHARED=-fPIC +endif + + +else + +LDFLAGS = $(LDFLAGS_STATIC) +# -s is not required for clang, do we need it for GGC ??? +# -s + +#-static -static-libgcc -static-libstdc++ + +ifdef IS_MINGW +SHARED_EXT=.exe +else +SHARED_EXT= +endif + +endif + + +PROGPATH = $(O)/$(PROG)$(SHARED_EXT) +PROGPATH_STATIC = $(O)/$(PROG)s$(SHARED_EXT) + +ifndef O +O=_o +endif + +ifdef IS_MINGW + +ifdef MSYSTEM +RM = rm -f +MY_MKDIR=mkdir -p +DEL_OBJ_EXE = -$(RM) $(PROGPATH) $(PROGPATH_STATIC) $(OBJS) +else +RM = del +MY_MKDIR=mkdir +DEL_OBJ_EXE = -$(RM) $(O)\*.o $(O)\$(PROG).exe $(O)\$(PROG).dll +endif + + +LIB2 = -lOle32 -loleaut32 -luuid -ladvapi32 -lUser32 + +CXXFLAGS_EXTRA = -DUNICODE -D_UNICODE +# -Wno-delete-non-virtual-dtor + + +else + +RM = rm -f +MY_MKDIR=mkdir -p +# CFLAGS_BASE := $(CFLAGS_BASE) -D_7ZIP_ST +# CXXFLAGS_EXTRA = -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE + +# LOCAL_LIBS=-lpthread +# LOCAL_LIBS_DLL=$(LOCAL_LIBS) -ldl +LIB2 = -lpthread -ldl + +DEL_OBJ_EXE = -$(RM) $(PROGPATH) $(PROGPATH_STATIC) $(OBJS) + +endif + + + +CFLAGS = $(LOCAL_FLAGS) $(CFLAGS_BASE2) $(CFLAGS_BASE) $(CC_SHARED) -o $@ + + +ifdef IS_X64 +AFLAGS_ABI = -elf64 -DABI_LINUX +else +AFLAGS_ABI = -elf -DABI_LINUX -DABI_CDECL +# -DABI_CDECL +# -DABI_LINUX +# -DABI_CDECL +endif +AFLAGS = $(AFLAGS_ABI) -Fo$(O)/ + + +CXX_WARN_FLAGS = +#-Wno-invalid-offsetof +#-Wno-reorder + +CXXFLAGS = $(LOCAL_FLAGS) $(CXXFLAGS_BASE2) $(CFLAGS_BASE) $(CXXFLAGS_EXTRA) $(CC_SHARED) -o $@ $(CXX_WARN_FLAGS) + +STATIC_TARGET= +ifdef COMPL_STATIC +STATIC_TARGET=$(PROGPATH_STATIC) +endif + + +all: $(O) $(PROGPATH) $(STATIC_TARGET) + +$(O): + $(MY_MKDIR) $(O) + +LFLAGS_ALL = -s $(MY_ARCH_2) $(LDFLAGS) $(LD_arch) $(OBJS) $(MY_LIBS) $(LIB2) +$(PROGPATH): $(OBJS) + $(CXX) -o $(PROGPATH) $(LFLAGS_ALL) + +$(PROGPATH_STATIC): $(OBJS) + $(CXX) -static -o $(PROGPATH_STATIC) $(LFLAGS_ALL) + + +ifndef NO_DEFAULT_RES +$O/resource.o: resource.rc + windres.exe $(RFLAGS) resource.rc $O/resource.o +endif + + + +$O/7zAlloc.o: ../../../C/7zAlloc.c + $(CC) $(CFLAGS) $< +$O/7zArcIn.o: ../../../C/7zArcIn.c + $(CC) $(CFLAGS) $< +$O/7zBuf.o: ../../../C/7zBuf.c + $(CC) $(CFLAGS) $< +$O/7zBuf2.o: ../../../C/7zBuf2.c + $(CC) $(CFLAGS) $< +$O/7zCrc.o: ../../../C/7zCrc.c + $(CC) $(CFLAGS) $< +$O/7zDec.o: ../../../C/7zDec.c + $(CC) $(CFLAGS) $< +$O/7zFile.o: ../../../C/7zFile.c + $(CC) $(CFLAGS) $< +$O/7zStream.o: ../../../C/7zStream.c + $(CC) $(CFLAGS) $< +$O/Aes.o: ../../../C/Aes.c + $(CC) $(CFLAGS) $< +$O/Alloc.o: ../../../C/Alloc.c + $(CC) $(CFLAGS) $< +$O/Bcj2.o: ../../../C/Bcj2.c + $(CC) $(CFLAGS) $< +$O/Bcj2Enc.o: ../../../C/Bcj2Enc.c + $(CC) $(CFLAGS) $< +$O/Blake2s.o: ../../../C/Blake2s.c + $(CC) $(CFLAGS) $< +$O/Bra.o: ../../../C/Bra.c + $(CC) $(CFLAGS) $< +$O/Bra86.o: ../../../C/Bra86.c + $(CC) $(CFLAGS) $< +$O/BraIA64.o: ../../../C/BraIA64.c + $(CC) $(CFLAGS) $< +$O/BwtSort.o: ../../../C/BwtSort.c + $(CC) $(CFLAGS) $< + +$O/CpuArch.o: ../../../C/CpuArch.c + $(CC) $(CFLAGS) $< +$O/Delta.o: ../../../C/Delta.c + $(CC) $(CFLAGS) $< +$O/DllSecur.o: ../../../C/DllSecur.c + $(CC) $(CFLAGS) $< +$O/HuffEnc.o: ../../../C/HuffEnc.c + $(CC) $(CFLAGS) $< +$O/LzFind.o: ../../../C/LzFind.c + $(CC) $(CFLAGS) $< + +# ifdef MT_FILES +$O/LzFindMt.o: ../../../C/LzFindMt.c + $(CC) $(CFLAGS) $< +$O/LzFindOpt.o: ../../../C/LzFindOpt.c + $(CC) $(CFLAGS) $< + +$O/Threads.o: ../../../C/Threads.c + $(CC) $(CFLAGS) $< +# endif + +$O/LzmaEnc.o: ../../../C/LzmaEnc.c + $(CC) $(CFLAGS) $< +$O/Lzma86Dec.o: ../../../C/Lzma86Dec.c + $(CC) $(CFLAGS) $< +$O/Lzma86Enc.o: ../../../C/Lzma86Enc.c + $(CC) $(CFLAGS) $< +$O/Lzma2Dec.o: ../../../C/Lzma2Dec.c + $(CC) $(CFLAGS) $< +$O/Lzma2DecMt.o: ../../../C/Lzma2DecMt.c + $(CC) $(CFLAGS) $< +$O/Lzma2Enc.o: ../../../C/Lzma2Enc.c + $(CC) $(CFLAGS) $< +$O/LzmaLib.o: ../../../C/LzmaLib.c + $(CC) $(CFLAGS) $< +$O/MtCoder.o: ../../../C/MtCoder.c + $(CC) $(CFLAGS) $< +$O/MtDec.o: ../../../C/MtDec.c + $(CC) $(CFLAGS) $< +$O/Ppmd7.o: ../../../C/Ppmd7.c + $(CC) $(CFLAGS) $< +$O/Ppmd7aDec.o: ../../../C/Ppmd7aDec.c + $(CC) $(CFLAGS) $< +$O/Ppmd7Dec.o: ../../../C/Ppmd7Dec.c + $(CC) $(CFLAGS) $< +$O/Ppmd7Enc.o: ../../../C/Ppmd7Enc.c + $(CC) $(CFLAGS) $< +$O/Ppmd8.o: ../../../C/Ppmd8.c + $(CC) $(CFLAGS) $< +$O/Ppmd8Dec.o: ../../../C/Ppmd8Dec.c + $(CC) $(CFLAGS) $< +$O/Ppmd8Enc.o: ../../../C/Ppmd8Enc.c + $(CC) $(CFLAGS) $< +$O/Sha1.o: ../../../C/Sha1.c + $(CC) $(CFLAGS) $< +$O/Sha256.o: ../../../C/Sha256.c + $(CC) $(CFLAGS) $< +$O/Sort.o: ../../../C/Sort.c + $(CC) $(CFLAGS) $< +$O/Xz.o: ../../../C/Xz.c + $(CC) $(CFLAGS) $< +$O/XzCrc64.o: ../../../C/XzCrc64.c + $(CC) $(CFLAGS) $< + + +ifdef USE_ASM +ifdef IS_X64 +USE_X86_ASM=1 +else +ifdef IS_X86 +USE_X86_ASM=1 +endif +endif +endif + +ifdef USE_X86_ASM +$O/7zCrcOpt.o: ../../../Asm/x86/7zCrcOpt.asm + $(MY_ASM) $(AFLAGS) $< +$O/XzCrc64Opt.o: ../../../Asm/x86/XzCrc64Opt.asm + $(MY_ASM) $(AFLAGS) $< +$O/AesOpt.o: ../../../Asm/x86/AesOpt.asm + $(MY_ASM) $(AFLAGS) $< +$O/Sha1Opt.o: ../../../Asm/x86/Sha1Opt.asm + $(MY_ASM) $(AFLAGS) $< +$O/Sha256Opt.o: ../../../Asm/x86/Sha256Opt.asm + $(MY_ASM) $(AFLAGS) $< +else +$O/7zCrcOpt.o: ../../7zCrcOpt.c + $(CC) $(CFLAGS) $< +$O/XzCrc64Opt.o: ../../XzCrc64Opt.c + $(CC) $(CFLAGS) $< +$O/Sha1Opt.o: ../../Sha1Opt.c + $(CC) $(CFLAGS) $< +$O/Sha256Opt.o: ../../Sha256Opt.c + $(CC) $(CFLAGS) $< +$O/AesOpt.o: ../../AesOpt.c + $(CC) $(CFLAGS) $< +endif + + +ifdef USE_LZMA_DEC_ASM + +ifdef IS_X64 +$O/LzmaDecOpt.o: ../../../Asm/x86/LzmaDecOpt.asm + $(MY_ASM) $(AFLAGS) $< +endif + +ifdef IS_ARM64 +$O/LzmaDecOpt.o: ../../../Asm/arm64/LzmaDecOpt.S ../../../Asm/arm64/7zAsm.S + $(CC) $(CFLAGS) $< +endif + +$O/LzmaDec.o: ../../LzmaDec.c + $(CC) $(CFLAGS) -D_LZMA_DEC_OPT $< + +else + +$O/LzmaDec.o: ../../LzmaDec.c + $(CC) $(CFLAGS) $< + +endif + + + +$O/XzDec.o: ../../../C/XzDec.c + $(CC) $(CFLAGS) $< +$O/XzEnc.o: ../../../C/XzEnc.c + $(CC) $(CFLAGS) $< +$O/XzIn.o: ../../../C/XzIn.c + $(CC) $(CFLAGS) $< + + +$O/7zMain.o: ../../../C/Util/7z/7zMain.c + $(CC) $(CFLAGS) $< +$O/LzmaUtil.o: ../../../C/Util/Lzma/LzmaUtil.c + $(CC) $(CFLAGS) $< +$O/7zipInstall.o: ../../../C/Util/7zipInstall/7zipInstall.c + $(CC) $(CFLAGS) $< +$O/7zipUninstall.o: ../../../C/Util/7zipUninstall/7zipUninstall.c + $(CC) $(CFLAGS) $< + + +clean: + -$(DEL_OBJ_EXE) diff --git a/deps/LZMA-SDK/C/Aes.c b/deps/LZMA-SDK/C/Aes.c index 9ad66c5c8..27e32e622 100644 --- a/deps/LZMA-SDK/C/Aes.c +++ b/deps/LZMA-SDK/C/Aes.c @@ -1,375 +1,375 @@ -/* Aes.c -- AES encryption / decryption -2021-05-13 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "CpuArch.h" -#include "Aes.h" - -AES_CODE_FUNC g_AesCbc_Decode; -#ifndef _SFX -AES_CODE_FUNC g_AesCbc_Encode; -AES_CODE_FUNC g_AesCtr_Code; -UInt32 g_Aes_SupportedFunctions_Flags; -#endif - -static UInt32 T[256 * 4]; -static const Byte Sbox[256] = { - 0x63, 0x7c, 0x77, 0x7b, 0xf2, 0x6b, 0x6f, 0xc5, 0x30, 0x01, 0x67, 0x2b, 0xfe, 0xd7, 0xab, 0x76, - 0xca, 0x82, 0xc9, 0x7d, 0xfa, 0x59, 0x47, 0xf0, 0xad, 0xd4, 0xa2, 0xaf, 0x9c, 0xa4, 0x72, 0xc0, - 0xb7, 0xfd, 0x93, 0x26, 0x36, 0x3f, 0xf7, 0xcc, 0x34, 0xa5, 0xe5, 0xf1, 0x71, 0xd8, 0x31, 0x15, - 0x04, 0xc7, 0x23, 0xc3, 0x18, 0x96, 0x05, 0x9a, 0x07, 0x12, 0x80, 0xe2, 0xeb, 0x27, 0xb2, 0x75, - 0x09, 0x83, 0x2c, 0x1a, 0x1b, 0x6e, 0x5a, 0xa0, 0x52, 0x3b, 0xd6, 0xb3, 0x29, 0xe3, 0x2f, 0x84, - 0x53, 0xd1, 0x00, 0xed, 0x20, 0xfc, 0xb1, 0x5b, 0x6a, 0xcb, 0xbe, 0x39, 0x4a, 0x4c, 0x58, 0xcf, - 0xd0, 0xef, 0xaa, 0xfb, 0x43, 0x4d, 0x33, 0x85, 0x45, 0xf9, 0x02, 0x7f, 0x50, 0x3c, 0x9f, 0xa8, - 0x51, 0xa3, 0x40, 0x8f, 0x92, 0x9d, 0x38, 0xf5, 0xbc, 0xb6, 0xda, 0x21, 0x10, 0xff, 0xf3, 0xd2, - 0xcd, 0x0c, 0x13, 0xec, 0x5f, 0x97, 0x44, 0x17, 0xc4, 0xa7, 0x7e, 0x3d, 0x64, 0x5d, 0x19, 0x73, - 0x60, 0x81, 0x4f, 0xdc, 0x22, 0x2a, 0x90, 0x88, 0x46, 0xee, 0xb8, 0x14, 0xde, 0x5e, 0x0b, 0xdb, - 0xe0, 0x32, 0x3a, 0x0a, 0x49, 0x06, 0x24, 0x5c, 0xc2, 0xd3, 0xac, 0x62, 0x91, 0x95, 0xe4, 0x79, - 0xe7, 0xc8, 0x37, 0x6d, 0x8d, 0xd5, 0x4e, 0xa9, 0x6c, 0x56, 0xf4, 0xea, 0x65, 0x7a, 0xae, 0x08, - 0xba, 0x78, 0x25, 0x2e, 0x1c, 0xa6, 0xb4, 0xc6, 0xe8, 0xdd, 0x74, 0x1f, 0x4b, 0xbd, 0x8b, 0x8a, - 0x70, 0x3e, 0xb5, 0x66, 0x48, 0x03, 0xf6, 0x0e, 0x61, 0x35, 0x57, 0xb9, 0x86, 0xc1, 0x1d, 0x9e, - 0xe1, 0xf8, 0x98, 0x11, 0x69, 0xd9, 0x8e, 0x94, 0x9b, 0x1e, 0x87, 0xe9, 0xce, 0x55, 0x28, 0xdf, - 0x8c, 0xa1, 0x89, 0x0d, 0xbf, 0xe6, 0x42, 0x68, 0x41, 0x99, 0x2d, 0x0f, 0xb0, 0x54, 0xbb, 0x16}; - - -static UInt32 D[256 * 4]; -static Byte InvS[256]; - -#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))) - -#define gb(n, x) gb ## n(x) - -#define TT(x) (T + (x << 8)) -#define DD(x) (D + (x << 8)) - - -// #define _SHOW_AES_STATUS - -#ifdef MY_CPU_X86_OR_AMD64 - #define USE_HW_AES -#elif defined(MY_CPU_ARM_OR_ARM64) && defined(MY_CPU_LE) - #if defined(__clang__) - #if (__clang_major__ >= 8) // fix that check - #define USE_HW_AES - #endif - #elif defined(__GNUC__) - #if (__GNUC__ >= 6) // fix that check - #define USE_HW_AES - #endif - #elif defined(_MSC_VER) - #if _MSC_VER >= 1910 - #define USE_HW_AES - #endif - #endif -#endif - -#ifdef USE_HW_AES -#ifdef _SHOW_AES_STATUS -#include -#define _PRF(x) x -#else -#define _PRF(x) -#endif -#endif - - -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; - TT(0)[i] = Ui32(a2, a1, a1, a3); - TT(1)[i] = Ui32(a3, a2, a1, a1); - TT(2)[i] = Ui32(a1, a3, a2, a1); - TT(3)[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; - DD(0)[i] = Ui32(aE, a9, aD, aB); - DD(1)[i] = Ui32(aB, aE, a9, aD); - DD(2)[i] = Ui32(aD, aB, aE, a9); - DD(3)[i] = Ui32(a9, aD, aB, aE); - } - } - - { - AES_CODE_FUNC d = AesCbc_Decode; - #ifndef _SFX - AES_CODE_FUNC e = AesCbc_Encode; - AES_CODE_FUNC c = AesCtr_Code; - UInt32 flags = 0; - #endif - - #ifdef USE_HW_AES - if (CPU_IsSupported_AES()) - { - // #pragma message ("AES HW") - _PRF(printf("\n===AES HW\n")); - d = AesCbc_Decode_HW; - - #ifndef _SFX - e = AesCbc_Encode_HW; - c = AesCtr_Code_HW; - flags = k_Aes_SupportedFunctions_HW; - #endif - - #ifdef MY_CPU_X86_OR_AMD64 - if (CPU_IsSupported_VAES_AVX2()) - { - _PRF(printf("\n===vaes avx2\n")); - d = AesCbc_Decode_HW_256; - #ifndef _SFX - c = AesCtr_Code_HW_256; - flags |= k_Aes_SupportedFunctions_HW_256; - #endif - } - #endif - } - #endif - - g_AesCbc_Decode = d; - #ifndef _SFX - g_AesCbc_Encode = e; - g_AesCtr_Code = c; - g_Aes_SupportedFunctions_Flags = flags; - #endif - } -} - - -#define HT(i, x, s) TT(x)[gb(x, s[(i + x) & 3])] - -#define HT4(m, i, s, p) m[i] = \ - HT(i, 0, s) ^ \ - HT(i, 1, s) ^ \ - HT(i, 2, s) ^ \ - HT(i, 3, s) ^ w[p + i] - -#define HT16(m, s, p) \ - HT4(m, 0, s, p); \ - HT4(m, 1, s, p); \ - HT4(m, 2, s, p); \ - HT4(m, 3, s, p); \ - -#define FT(i, x) Sbox[gb(x, m[(i + x) & 3])] -#define FT4(i) dest[i] = Ui32(FT(i, 0), FT(i, 1), FT(i, 2), FT(i, 3)) ^ w[i]; - - -#define HD(i, x, s) DD(x)[gb(x, s[(i - x) & 3])] - -#define HD4(m, i, s, p) m[i] = \ - HD(i, 0, s) ^ \ - HD(i, 1, s) ^ \ - HD(i, 2, s) ^ \ - HD(i, 3, s) ^ w[p + i]; - -#define HD16(m, s, p) \ - HD4(m, 0, s, p); \ - HD4(m, 1, s, p); \ - HD4(m, 2, s, p); \ - HD4(m, 3, s, p); \ - -#define FD(i, x) InvS[gb(x, m[(i - x) & 3])] -#define FD4(i) dest[i] = Ui32(FD(i, 0), FD(i, 1), FD(i, 2), FD(i, 3)) ^ w[i]; - -void MY_FAST_CALL Aes_SetKey_Enc(UInt32 *w, const Byte *key, unsigned keySize) -{ - unsigned i, m; - const UInt32 *wLim; - UInt32 t; - UInt32 rcon = 1; - - keySize /= 4; - w[0] = ((UInt32)keySize / 2) + 3; - w += 4; - - for (i = 0; i < keySize; i++, key += 4) - w[i] = GetUi32(key); - - t = w[(size_t)keySize - 1]; - wLim = w + (size_t)keySize * 3 + 28; - m = 0; - do - { - if (m == 0) - { - t = Ui32(Sbox[gb1(t)] ^ rcon, Sbox[gb2(t)], Sbox[gb3(t)], Sbox[gb0(t)]); - rcon <<= 1; - if (rcon & 0x100) - rcon = 0x1b; - m = keySize; - } - else if (m == 4 && keySize > 6) - t = Ui32(Sbox[gb0(t)], Sbox[gb1(t)], Sbox[gb2(t)], Sbox[gb3(t)]); - m--; - t ^= w[0]; - w[keySize] = t; - } - while (++w != wLim); -} - -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] = - DD(0)[Sbox[gb0(r)]] ^ - DD(1)[Sbox[gb1(r)]] ^ - DD(2)[Sbox[gb2(r)]] ^ - DD(3)[Sbox[gb3(r)]]; - } -} - -/* Aes_Encode and Aes_Decode functions work with little-endian words. - src and dest are pointers to 4 UInt32 words. - src and dest can point to same block */ - -// MY_FORCE_INLINE -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); -} - -MY_FORCE_INLINE -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]; - unsigned i; - - if (++p[0] == 0) - p[1]++; - - Aes_Encode(p + 4, temp, p); - - for (i = 0; i < 4; i++, data += 4) - { - UInt32 t = temp[i]; - - #ifdef MY_CPU_LE_UNALIGN - *((UInt32 *)(void *)data) ^= t; - #else - data[0] = (Byte)(data[0] ^ (t & 0xFF)); - data[1] = (Byte)(data[1] ^ ((t >> 8) & 0xFF)); - data[2] = (Byte)(data[2] ^ ((t >> 16) & 0xFF)); - data[3] = (Byte)(data[3] ^ ((t >> 24))); - #endif - } - } -} +/* Aes.c -- AES encryption / decryption +2021-05-13 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" +#include "Aes.h" + +AES_CODE_FUNC g_AesCbc_Decode; +#ifndef _SFX +AES_CODE_FUNC g_AesCbc_Encode; +AES_CODE_FUNC g_AesCtr_Code; +UInt32 g_Aes_SupportedFunctions_Flags; +#endif + +static UInt32 T[256 * 4]; +static const Byte Sbox[256] = { + 0x63, 0x7c, 0x77, 0x7b, 0xf2, 0x6b, 0x6f, 0xc5, 0x30, 0x01, 0x67, 0x2b, 0xfe, 0xd7, 0xab, 0x76, + 0xca, 0x82, 0xc9, 0x7d, 0xfa, 0x59, 0x47, 0xf0, 0xad, 0xd4, 0xa2, 0xaf, 0x9c, 0xa4, 0x72, 0xc0, + 0xb7, 0xfd, 0x93, 0x26, 0x36, 0x3f, 0xf7, 0xcc, 0x34, 0xa5, 0xe5, 0xf1, 0x71, 0xd8, 0x31, 0x15, + 0x04, 0xc7, 0x23, 0xc3, 0x18, 0x96, 0x05, 0x9a, 0x07, 0x12, 0x80, 0xe2, 0xeb, 0x27, 0xb2, 0x75, + 0x09, 0x83, 0x2c, 0x1a, 0x1b, 0x6e, 0x5a, 0xa0, 0x52, 0x3b, 0xd6, 0xb3, 0x29, 0xe3, 0x2f, 0x84, + 0x53, 0xd1, 0x00, 0xed, 0x20, 0xfc, 0xb1, 0x5b, 0x6a, 0xcb, 0xbe, 0x39, 0x4a, 0x4c, 0x58, 0xcf, + 0xd0, 0xef, 0xaa, 0xfb, 0x43, 0x4d, 0x33, 0x85, 0x45, 0xf9, 0x02, 0x7f, 0x50, 0x3c, 0x9f, 0xa8, + 0x51, 0xa3, 0x40, 0x8f, 0x92, 0x9d, 0x38, 0xf5, 0xbc, 0xb6, 0xda, 0x21, 0x10, 0xff, 0xf3, 0xd2, + 0xcd, 0x0c, 0x13, 0xec, 0x5f, 0x97, 0x44, 0x17, 0xc4, 0xa7, 0x7e, 0x3d, 0x64, 0x5d, 0x19, 0x73, + 0x60, 0x81, 0x4f, 0xdc, 0x22, 0x2a, 0x90, 0x88, 0x46, 0xee, 0xb8, 0x14, 0xde, 0x5e, 0x0b, 0xdb, + 0xe0, 0x32, 0x3a, 0x0a, 0x49, 0x06, 0x24, 0x5c, 0xc2, 0xd3, 0xac, 0x62, 0x91, 0x95, 0xe4, 0x79, + 0xe7, 0xc8, 0x37, 0x6d, 0x8d, 0xd5, 0x4e, 0xa9, 0x6c, 0x56, 0xf4, 0xea, 0x65, 0x7a, 0xae, 0x08, + 0xba, 0x78, 0x25, 0x2e, 0x1c, 0xa6, 0xb4, 0xc6, 0xe8, 0xdd, 0x74, 0x1f, 0x4b, 0xbd, 0x8b, 0x8a, + 0x70, 0x3e, 0xb5, 0x66, 0x48, 0x03, 0xf6, 0x0e, 0x61, 0x35, 0x57, 0xb9, 0x86, 0xc1, 0x1d, 0x9e, + 0xe1, 0xf8, 0x98, 0x11, 0x69, 0xd9, 0x8e, 0x94, 0x9b, 0x1e, 0x87, 0xe9, 0xce, 0x55, 0x28, 0xdf, + 0x8c, 0xa1, 0x89, 0x0d, 0xbf, 0xe6, 0x42, 0x68, 0x41, 0x99, 0x2d, 0x0f, 0xb0, 0x54, 0xbb, 0x16}; + + +static UInt32 D[256 * 4]; +static Byte InvS[256]; + +#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))) + +#define gb(n, x) gb ## n(x) + +#define TT(x) (T + (x << 8)) +#define DD(x) (D + (x << 8)) + + +// #define _SHOW_AES_STATUS + +#ifdef MY_CPU_X86_OR_AMD64 + #define USE_HW_AES +#elif defined(MY_CPU_ARM_OR_ARM64) && defined(MY_CPU_LE) + #if defined(__clang__) + #if (__clang_major__ >= 8) // fix that check + #define USE_HW_AES + #endif + #elif defined(__GNUC__) + #if (__GNUC__ >= 6) // fix that check + #define USE_HW_AES + #endif + #elif defined(_MSC_VER) + #if _MSC_VER >= 1910 + #define USE_HW_AES + #endif + #endif +#endif + +#ifdef USE_HW_AES +#ifdef _SHOW_AES_STATUS +#include +#define _PRF(x) x +#else +#define _PRF(x) +#endif +#endif + + +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; + TT(0)[i] = Ui32(a2, a1, a1, a3); + TT(1)[i] = Ui32(a3, a2, a1, a1); + TT(2)[i] = Ui32(a1, a3, a2, a1); + TT(3)[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; + DD(0)[i] = Ui32(aE, a9, aD, aB); + DD(1)[i] = Ui32(aB, aE, a9, aD); + DD(2)[i] = Ui32(aD, aB, aE, a9); + DD(3)[i] = Ui32(a9, aD, aB, aE); + } + } + + { + AES_CODE_FUNC d = AesCbc_Decode; + #ifndef _SFX + AES_CODE_FUNC e = AesCbc_Encode; + AES_CODE_FUNC c = AesCtr_Code; + UInt32 flags = 0; + #endif + + #ifdef USE_HW_AES + if (CPU_IsSupported_AES()) + { + // #pragma message ("AES HW") + _PRF(printf("\n===AES HW\n")); + d = AesCbc_Decode_HW; + + #ifndef _SFX + e = AesCbc_Encode_HW; + c = AesCtr_Code_HW; + flags = k_Aes_SupportedFunctions_HW; + #endif + + #ifdef MY_CPU_X86_OR_AMD64 + if (CPU_IsSupported_VAES_AVX2()) + { + _PRF(printf("\n===vaes avx2\n")); + d = AesCbc_Decode_HW_256; + #ifndef _SFX + c = AesCtr_Code_HW_256; + flags |= k_Aes_SupportedFunctions_HW_256; + #endif + } + #endif + } + #endif + + g_AesCbc_Decode = d; + #ifndef _SFX + g_AesCbc_Encode = e; + g_AesCtr_Code = c; + g_Aes_SupportedFunctions_Flags = flags; + #endif + } +} + + +#define HT(i, x, s) TT(x)[gb(x, s[(i + x) & 3])] + +#define HT4(m, i, s, p) m[i] = \ + HT(i, 0, s) ^ \ + HT(i, 1, s) ^ \ + HT(i, 2, s) ^ \ + HT(i, 3, s) ^ w[p + i] + +#define HT16(m, s, p) \ + HT4(m, 0, s, p); \ + HT4(m, 1, s, p); \ + HT4(m, 2, s, p); \ + HT4(m, 3, s, p); \ + +#define FT(i, x) Sbox[gb(x, m[(i + x) & 3])] +#define FT4(i) dest[i] = Ui32(FT(i, 0), FT(i, 1), FT(i, 2), FT(i, 3)) ^ w[i]; + + +#define HD(i, x, s) DD(x)[gb(x, s[(i - x) & 3])] + +#define HD4(m, i, s, p) m[i] = \ + HD(i, 0, s) ^ \ + HD(i, 1, s) ^ \ + HD(i, 2, s) ^ \ + HD(i, 3, s) ^ w[p + i]; + +#define HD16(m, s, p) \ + HD4(m, 0, s, p); \ + HD4(m, 1, s, p); \ + HD4(m, 2, s, p); \ + HD4(m, 3, s, p); \ + +#define FD(i, x) InvS[gb(x, m[(i - x) & 3])] +#define FD4(i) dest[i] = Ui32(FD(i, 0), FD(i, 1), FD(i, 2), FD(i, 3)) ^ w[i]; + +void MY_FAST_CALL Aes_SetKey_Enc(UInt32 *w, const Byte *key, unsigned keySize) +{ + unsigned i, m; + const UInt32 *wLim; + UInt32 t; + UInt32 rcon = 1; + + keySize /= 4; + w[0] = ((UInt32)keySize / 2) + 3; + w += 4; + + for (i = 0; i < keySize; i++, key += 4) + w[i] = GetUi32(key); + + t = w[(size_t)keySize - 1]; + wLim = w + (size_t)keySize * 3 + 28; + m = 0; + do + { + if (m == 0) + { + t = Ui32(Sbox[gb1(t)] ^ rcon, Sbox[gb2(t)], Sbox[gb3(t)], Sbox[gb0(t)]); + rcon <<= 1; + if (rcon & 0x100) + rcon = 0x1b; + m = keySize; + } + else if (m == 4 && keySize > 6) + t = Ui32(Sbox[gb0(t)], Sbox[gb1(t)], Sbox[gb2(t)], Sbox[gb3(t)]); + m--; + t ^= w[0]; + w[keySize] = t; + } + while (++w != wLim); +} + +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] = + DD(0)[Sbox[gb0(r)]] ^ + DD(1)[Sbox[gb1(r)]] ^ + DD(2)[Sbox[gb2(r)]] ^ + DD(3)[Sbox[gb3(r)]]; + } +} + +/* Aes_Encode and Aes_Decode functions work with little-endian words. + src and dest are pointers to 4 UInt32 words. + src and dest can point to same block */ + +// MY_FORCE_INLINE +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); +} + +MY_FORCE_INLINE +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]; + unsigned i; + + if (++p[0] == 0) + p[1]++; + + Aes_Encode(p + 4, temp, p); + + for (i = 0; i < 4; i++, data += 4) + { + UInt32 t = temp[i]; + + #ifdef MY_CPU_LE_UNALIGN + *((UInt32 *)(void *)data) ^= t; + #else + data[0] = (Byte)(data[0] ^ (t & 0xFF)); + data[1] = (Byte)(data[1] ^ ((t >> 8) & 0xFF)); + data[2] = (Byte)(data[2] ^ ((t >> 16) & 0xFF)); + data[3] = (Byte)(data[3] ^ ((t >> 24))); + #endif + } + } +} diff --git a/deps/LZMA-SDK/C/Aes.h b/deps/LZMA-SDK/C/Aes.h index 602e25ea2..2aa225642 100644 --- a/deps/LZMA-SDK/C/Aes.h +++ b/deps/LZMA-SDK/C/Aes.h @@ -1,60 +1,60 @@ -/* Aes.h -- AES encryption / decryption -2018-04-28 : Igor Pavlov : Public domain */ - -#ifndef __AES_H -#define __AES_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -#define AES_BLOCK_SIZE 16 - -/* Call AesGenTables one time before other AES functions */ -void AesGenTables(void); - -/* UInt32 pointers must be 16-byte aligned */ - -/* 16-byte (4 * 32-bit words) blocks: 1 (IV) + 1 (keyMode) + 15 (AES-256 roundKeys) */ -#define AES_NUM_IVMRK_WORDS ((1 + 1 + 15) * 4) - -/* aes - 16-byte aligned pointer to keyMode+roundKeys sequence */ -/* keySize = 16 or 24 or 32 (bytes) */ -typedef void (MY_FAST_CALL *AES_SET_KEY_FUNC)(UInt32 *aes, const Byte *key, unsigned keySize); -void MY_FAST_CALL Aes_SetKey_Enc(UInt32 *aes, const Byte *key, unsigned keySize); -void MY_FAST_CALL Aes_SetKey_Dec(UInt32 *aes, const Byte *key, unsigned keySize); - -/* ivAes - 16-byte aligned pointer to iv+keyMode+roundKeys sequence: UInt32[AES_NUM_IVMRK_WORDS] */ -void AesCbc_Init(UInt32 *ivAes, const Byte *iv); /* iv size is AES_BLOCK_SIZE */ - -/* data - 16-byte aligned pointer to data */ -/* numBlocks - the number of 16-byte blocks in data array */ -typedef void (MY_FAST_CALL *AES_CODE_FUNC)(UInt32 *ivAes, Byte *data, size_t numBlocks); - -extern AES_CODE_FUNC g_AesCbc_Decode; -#ifndef _SFX -extern AES_CODE_FUNC g_AesCbc_Encode; -extern AES_CODE_FUNC g_AesCtr_Code; -#define k_Aes_SupportedFunctions_HW (1 << 2) -#define k_Aes_SupportedFunctions_HW_256 (1 << 3) -extern UInt32 g_Aes_SupportedFunctions_Flags; -#endif - - -#define DECLARE__AES_CODE_FUNC(funcName) \ - void MY_FAST_CALL funcName(UInt32 *ivAes, Byte *data, size_t numBlocks); - -DECLARE__AES_CODE_FUNC (AesCbc_Encode) -DECLARE__AES_CODE_FUNC (AesCbc_Decode) -DECLARE__AES_CODE_FUNC (AesCtr_Code) - -DECLARE__AES_CODE_FUNC (AesCbc_Encode_HW) -DECLARE__AES_CODE_FUNC (AesCbc_Decode_HW) -DECLARE__AES_CODE_FUNC (AesCtr_Code_HW) - -DECLARE__AES_CODE_FUNC (AesCbc_Decode_HW_256) -DECLARE__AES_CODE_FUNC (AesCtr_Code_HW_256) - -EXTERN_C_END - -#endif +/* Aes.h -- AES encryption / decryption +2018-04-28 : Igor Pavlov : Public domain */ + +#ifndef __AES_H +#define __AES_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define AES_BLOCK_SIZE 16 + +/* Call AesGenTables one time before other AES functions */ +void AesGenTables(void); + +/* UInt32 pointers must be 16-byte aligned */ + +/* 16-byte (4 * 32-bit words) blocks: 1 (IV) + 1 (keyMode) + 15 (AES-256 roundKeys) */ +#define AES_NUM_IVMRK_WORDS ((1 + 1 + 15) * 4) + +/* aes - 16-byte aligned pointer to keyMode+roundKeys sequence */ +/* keySize = 16 or 24 or 32 (bytes) */ +typedef void (MY_FAST_CALL *AES_SET_KEY_FUNC)(UInt32 *aes, const Byte *key, unsigned keySize); +void MY_FAST_CALL Aes_SetKey_Enc(UInt32 *aes, const Byte *key, unsigned keySize); +void MY_FAST_CALL Aes_SetKey_Dec(UInt32 *aes, const Byte *key, unsigned keySize); + +/* ivAes - 16-byte aligned pointer to iv+keyMode+roundKeys sequence: UInt32[AES_NUM_IVMRK_WORDS] */ +void AesCbc_Init(UInt32 *ivAes, const Byte *iv); /* iv size is AES_BLOCK_SIZE */ + +/* data - 16-byte aligned pointer to data */ +/* numBlocks - the number of 16-byte blocks in data array */ +typedef void (MY_FAST_CALL *AES_CODE_FUNC)(UInt32 *ivAes, Byte *data, size_t numBlocks); + +extern AES_CODE_FUNC g_AesCbc_Decode; +#ifndef _SFX +extern AES_CODE_FUNC g_AesCbc_Encode; +extern AES_CODE_FUNC g_AesCtr_Code; +#define k_Aes_SupportedFunctions_HW (1 << 2) +#define k_Aes_SupportedFunctions_HW_256 (1 << 3) +extern UInt32 g_Aes_SupportedFunctions_Flags; +#endif + + +#define DECLARE__AES_CODE_FUNC(funcName) \ + void MY_FAST_CALL funcName(UInt32 *ivAes, Byte *data, size_t numBlocks); + +DECLARE__AES_CODE_FUNC (AesCbc_Encode) +DECLARE__AES_CODE_FUNC (AesCbc_Decode) +DECLARE__AES_CODE_FUNC (AesCtr_Code) + +DECLARE__AES_CODE_FUNC (AesCbc_Encode_HW) +DECLARE__AES_CODE_FUNC (AesCbc_Decode_HW) +DECLARE__AES_CODE_FUNC (AesCtr_Code_HW) + +DECLARE__AES_CODE_FUNC (AesCbc_Decode_HW_256) +DECLARE__AES_CODE_FUNC (AesCtr_Code_HW_256) + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/AesOpt.c b/deps/LZMA-SDK/C/AesOpt.c index 1bdc9a882..8be8ff69d 100644 --- a/deps/LZMA-SDK/C/AesOpt.c +++ b/deps/LZMA-SDK/C/AesOpt.c @@ -1,776 +1,776 @@ -/* AesOpt.c -- AES optimized code for x86 AES hardware instructions -2021-04-01 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "CpuArch.h" - -#ifdef MY_CPU_X86_OR_AMD64 - - #if defined(__clang__) - #if __clang_major__ > 3 || (__clang_major__ == 3 && __clang_minor__ >= 8) - #define USE_INTEL_AES - #define ATTRIB_AES __attribute__((__target__("aes"))) - #if (__clang_major__ >= 8) - #define USE_INTEL_VAES - #define ATTRIB_VAES __attribute__((__target__("aes,vaes,avx2"))) - #endif - #endif - #elif defined(__GNUC__) - #if (__GNUC__ > 4) || (__GNUC__ == 4 && __GNUC_MINOR__ >= 4) - #define USE_INTEL_AES - #ifndef __AES__ - #define ATTRIB_AES __attribute__((__target__("aes"))) - #endif - #if (__GNUC__ >= 8) - #define USE_INTEL_VAES - #define ATTRIB_VAES __attribute__((__target__("aes,vaes,avx2"))) - #endif - #endif - #elif defined(__INTEL_COMPILER) - #if (__INTEL_COMPILER >= 1110) - #define USE_INTEL_AES - #if (__INTEL_COMPILER >= 1900) - #define USE_INTEL_VAES - #endif - #endif - #elif defined(_MSC_VER) - #if (_MSC_VER > 1500) || (_MSC_FULL_VER >= 150030729) - #define USE_INTEL_AES - #if (_MSC_VER >= 1910) - #define USE_INTEL_VAES - #endif - #endif - #endif - -#ifndef ATTRIB_AES - #define ATTRIB_AES -#endif -#ifndef ATTRIB_VAES - #define ATTRIB_VAES -#endif - - -#ifdef USE_INTEL_AES - -#include - -#ifndef USE_INTEL_VAES -#define AES_TYPE_keys __m128i -#define AES_TYPE_data __m128i -#endif - -#define AES_FUNC_START(name) \ - void MY_FAST_CALL name(__m128i *p, __m128i *data, size_t numBlocks) - -#define AES_FUNC_START2(name) \ -AES_FUNC_START (name); \ -ATTRIB_AES \ -AES_FUNC_START (name) - -#define MM_OP(op, dest, src) dest = op(dest, src); -#define MM_OP_m(op, src) MM_OP(op, m, src); - -#define MM_XOR( dest, src) MM_OP(_mm_xor_si128, dest, src); -#define AVX_XOR(dest, src) MM_OP(_mm256_xor_si256, dest, src); - - -AES_FUNC_START2 (AesCbc_Encode_HW) -{ - __m128i m = *p; - const __m128i k0 = p[2]; - const __m128i k1 = p[3]; - const UInt32 numRounds2 = *(const UInt32 *)(p + 1) - 1; - for (; numBlocks != 0; numBlocks--, data++) - { - UInt32 r = numRounds2; - const __m128i *w = p + 4; - __m128i temp = *data; - MM_XOR (temp, k0); - MM_XOR (m, temp); - MM_OP_m (_mm_aesenc_si128, k1); - do - { - MM_OP_m (_mm_aesenc_si128, w[0]); - MM_OP_m (_mm_aesenc_si128, w[1]); - w += 2; - } - while (--r); - MM_OP_m (_mm_aesenclast_si128, w[0]); - *data = m; - } - *p = m; -} - - -#define WOP_1(op) -#define WOP_2(op) WOP_1 (op) op (m1, 1); -#define WOP_3(op) WOP_2 (op) op (m2, 2); -#define WOP_4(op) WOP_3 (op) op (m3, 3); -#ifdef MY_CPU_AMD64 -#define WOP_5(op) WOP_4 (op) op (m4, 4); -#define WOP_6(op) WOP_5 (op) op (m5, 5); -#define WOP_7(op) WOP_6 (op) op (m6, 6); -#define WOP_8(op) WOP_7 (op) op (m7, 7); -#endif -/* -#define WOP_9(op) WOP_8 (op) op (m8, 8); -#define WOP_10(op) WOP_9 (op) op (m9, 9); -#define WOP_11(op) WOP_10(op) op (m10, 10); -#define WOP_12(op) WOP_11(op) op (m11, 11); -#define WOP_13(op) WOP_12(op) op (m12, 12); -#define WOP_14(op) WOP_13(op) op (m13, 13); -*/ - -#ifdef MY_CPU_AMD64 - #define NUM_WAYS 8 - #define WOP_M1 WOP_8 -#else - #define NUM_WAYS 4 - #define WOP_M1 WOP_4 -#endif - -#define WOP(op) op (m0, 0); WOP_M1(op) - - -#define DECLARE_VAR(reg, ii) __m128i reg -#define LOAD_data( reg, ii) reg = data[ii]; -#define STORE_data( reg, ii) data[ii] = reg; -#if (NUM_WAYS > 1) -#define XOR_data_M1(reg, ii) MM_XOR (reg, data[ii- 1]); -#endif - -#define AVX__DECLARE_VAR(reg, ii) __m256i reg -#define AVX__LOAD_data( reg, ii) reg = ((const __m256i *)(const void *)data)[ii]; -#define AVX__STORE_data( reg, ii) ((__m256i *)(void *)data)[ii] = reg; -#define AVX__XOR_data_M1(reg, ii) AVX_XOR (reg, (((const __m256i *)(const void *)(data - 1))[ii])); - -#define MM_OP_key(op, reg) MM_OP(op, reg, key); - -#define AES_DEC( reg, ii) MM_OP_key (_mm_aesdec_si128, reg) -#define AES_DEC_LAST( reg, ii) MM_OP_key (_mm_aesdeclast_si128, reg) -#define AES_ENC( reg, ii) MM_OP_key (_mm_aesenc_si128, reg) -#define AES_ENC_LAST( reg, ii) MM_OP_key (_mm_aesenclast_si128, reg) -#define AES_XOR( reg, ii) MM_OP_key (_mm_xor_si128, reg) - - -#define AVX__AES_DEC( reg, ii) MM_OP_key (_mm256_aesdec_epi128, reg) -#define AVX__AES_DEC_LAST( reg, ii) MM_OP_key (_mm256_aesdeclast_epi128, reg) -#define AVX__AES_ENC( reg, ii) MM_OP_key (_mm256_aesenc_epi128, reg) -#define AVX__AES_ENC_LAST( reg, ii) MM_OP_key (_mm256_aesenclast_epi128, reg) -#define AVX__AES_XOR( reg, ii) MM_OP_key (_mm256_xor_si256, reg) - -#define CTR_START(reg, ii) MM_OP (_mm_add_epi64, ctr, one); reg = ctr; -#define CTR_END( reg, ii) MM_XOR (data[ii], reg); - -#define AVX__CTR_START(reg, ii) MM_OP (_mm256_add_epi64, ctr2, two); reg = _mm256_xor_si256(ctr2, key); -#define AVX__CTR_END( reg, ii) AVX_XOR (((__m256i *)(void *)data)[ii], reg); - -#define WOP_KEY(op, n) { \ - const __m128i key = w[n]; \ - WOP(op); } - -#define AVX__WOP_KEY(op, n) { \ - const __m256i key = w[n]; \ - WOP(op); } - - -#define WIDE_LOOP_START \ - dataEnd = data + numBlocks; \ - if (numBlocks >= NUM_WAYS) \ - { dataEnd -= NUM_WAYS; do { \ - - -#define WIDE_LOOP_END \ - data += NUM_WAYS; \ - } while (data <= dataEnd); \ - dataEnd += NUM_WAYS; } \ - - -#define SINGLE_LOOP \ - for (; data < dataEnd; data++) - - -#define NUM_AES_KEYS_MAX 15 - -#define WIDE_LOOP_START_AVX(OP) \ - dataEnd = data + numBlocks; \ - if (numBlocks >= NUM_WAYS * 2) \ - { __m256i keys[NUM_AES_KEYS_MAX]; \ - UInt32 ii; \ - OP \ - for (ii = 0; ii < numRounds; ii++) \ - keys[ii] = _mm256_broadcastsi128_si256(p[ii]); \ - dataEnd -= NUM_WAYS * 2; do { \ - - -#define WIDE_LOOP_END_AVX(OP) \ - data += NUM_WAYS * 2; \ - } while (data <= dataEnd); \ - dataEnd += NUM_WAYS * 2; \ - OP \ - _mm256_zeroupper(); \ - } \ - -/* MSVC for x86: If we don't call _mm256_zeroupper(), and -arch:IA32 is not specified, - MSVC still can insert vzeroupper instruction. */ - - -AES_FUNC_START2 (AesCbc_Decode_HW) -{ - __m128i iv = *p; - const __m128i *wStart = p + *(const UInt32 *)(p + 1) * 2 + 2 - 1; - const __m128i *dataEnd; - p += 2; - - WIDE_LOOP_START - { - const __m128i *w = wStart; - - WOP (DECLARE_VAR) - WOP (LOAD_data); - WOP_KEY (AES_XOR, 1) - - do - { - WOP_KEY (AES_DEC, 0) - w--; - } - while (w != p); - WOP_KEY (AES_DEC_LAST, 0) - - MM_XOR (m0, iv); - WOP_M1 (XOR_data_M1) - iv = data[NUM_WAYS - 1]; - WOP (STORE_data); - } - WIDE_LOOP_END - - SINGLE_LOOP - { - const __m128i *w = wStart - 1; - __m128i m = _mm_xor_si128 (w[2], *data); - do - { - MM_OP_m (_mm_aesdec_si128, w[1]); - MM_OP_m (_mm_aesdec_si128, w[0]); - w -= 2; - } - while (w != p); - MM_OP_m (_mm_aesdec_si128, w[1]); - MM_OP_m (_mm_aesdeclast_si128, w[0]); - - MM_XOR (m, iv); - iv = *data; - *data = m; - } - - p[-2] = iv; -} - - -AES_FUNC_START2 (AesCtr_Code_HW) -{ - __m128i ctr = *p; - UInt32 numRoundsMinus2 = *(const UInt32 *)(p + 1) * 2 - 1; - const __m128i *dataEnd; - __m128i one = _mm_cvtsi32_si128(1); - - p += 2; - - WIDE_LOOP_START - { - const __m128i *w = p; - UInt32 r = numRoundsMinus2; - WOP (DECLARE_VAR) - WOP (CTR_START); - WOP_KEY (AES_XOR, 0) - w += 1; - do - { - WOP_KEY (AES_ENC, 0) - w += 1; - } - while (--r); - WOP_KEY (AES_ENC_LAST, 0) - - WOP (CTR_END); - } - WIDE_LOOP_END - - SINGLE_LOOP - { - UInt32 numRounds2 = *(const UInt32 *)(p - 2 + 1) - 1; - const __m128i *w = p; - __m128i m; - MM_OP (_mm_add_epi64, ctr, one); - m = _mm_xor_si128 (ctr, p[0]); - w += 1; - do - { - MM_OP_m (_mm_aesenc_si128, w[0]); - MM_OP_m (_mm_aesenc_si128, w[1]); - w += 2; - } - while (--numRounds2); - MM_OP_m (_mm_aesenc_si128, w[0]); - MM_OP_m (_mm_aesenclast_si128, w[1]); - MM_XOR (*data, m); - } - - p[-2] = ctr; -} - - - -#ifdef USE_INTEL_VAES - -#if defined(__clang__) && defined(_MSC_VER) -#define __SSE4_2__ -#define __AES__ -#define __AVX__ -#define __AVX2__ -#define __VAES__ -#define __AVX512F__ -#define __AVX512VL__ -#endif - -#include - -#define VAES_FUNC_START2(name) \ -AES_FUNC_START (name); \ -ATTRIB_VAES \ -AES_FUNC_START (name) - -VAES_FUNC_START2 (AesCbc_Decode_HW_256) -{ - __m128i iv = *p; - const __m128i *dataEnd; - UInt32 numRounds = *(const UInt32 *)(p + 1) * 2 + 1; - p += 2; - - WIDE_LOOP_START_AVX(;) - { - const __m256i *w = keys + numRounds - 2; - - WOP (AVX__DECLARE_VAR) - WOP (AVX__LOAD_data); - AVX__WOP_KEY (AVX__AES_XOR, 1) - - do - { - AVX__WOP_KEY (AVX__AES_DEC, 0) - w--; - } - while (w != keys); - AVX__WOP_KEY (AVX__AES_DEC_LAST, 0) - - AVX_XOR (m0, _mm256_setr_m128i(iv, data[0])); - WOP_M1 (AVX__XOR_data_M1) - iv = data[NUM_WAYS * 2 - 1]; - WOP (AVX__STORE_data); - } - WIDE_LOOP_END_AVX(;) - - SINGLE_LOOP - { - const __m128i *w = p + *(const UInt32 *)(p + 1 - 2) * 2 + 1 - 3; - __m128i m = _mm_xor_si128 (w[2], *data); - do - { - MM_OP_m (_mm_aesdec_si128, w[1]); - MM_OP_m (_mm_aesdec_si128, w[0]); - w -= 2; - } - while (w != p); - MM_OP_m (_mm_aesdec_si128, w[1]); - MM_OP_m (_mm_aesdeclast_si128, w[0]); - - MM_XOR (m, iv); - iv = *data; - *data = m; - } - - p[-2] = iv; -} - - -/* -SSE2: _mm_cvtsi32_si128 : movd -AVX: _mm256_setr_m128i : vinsertf128 -AVX2: _mm256_add_epi64 : vpaddq ymm, ymm, ymm - _mm256_extracti128_si256 : vextracti128 - _mm256_broadcastsi128_si256 : vbroadcasti128 -*/ - -#define AVX__CTR_LOOP_START \ - ctr2 = _mm256_setr_m128i(_mm_sub_epi64(ctr, one), ctr); \ - two = _mm256_setr_m128i(one, one); \ - two = _mm256_add_epi64(two, two); \ - -// two = _mm256_setr_epi64x(2, 0, 2, 0); - -#define AVX__CTR_LOOP_ENC \ - ctr = _mm256_extracti128_si256 (ctr2, 1); \ - -VAES_FUNC_START2 (AesCtr_Code_HW_256) -{ - __m128i ctr = *p; - UInt32 numRounds = *(const UInt32 *)(p + 1) * 2 + 1; - const __m128i *dataEnd; - __m128i one = _mm_cvtsi32_si128(1); - __m256i ctr2, two; - p += 2; - - WIDE_LOOP_START_AVX (AVX__CTR_LOOP_START) - { - const __m256i *w = keys; - UInt32 r = numRounds - 2; - WOP (AVX__DECLARE_VAR) - AVX__WOP_KEY (AVX__CTR_START, 0); - - w += 1; - do - { - AVX__WOP_KEY (AVX__AES_ENC, 0) - w += 1; - } - while (--r); - AVX__WOP_KEY (AVX__AES_ENC_LAST, 0) - - WOP (AVX__CTR_END); - } - WIDE_LOOP_END_AVX (AVX__CTR_LOOP_ENC) - - SINGLE_LOOP - { - UInt32 numRounds2 = *(const UInt32 *)(p - 2 + 1) - 1; - const __m128i *w = p; - __m128i m; - MM_OP (_mm_add_epi64, ctr, one); - m = _mm_xor_si128 (ctr, p[0]); - w += 1; - do - { - MM_OP_m (_mm_aesenc_si128, w[0]); - MM_OP_m (_mm_aesenc_si128, w[1]); - w += 2; - } - while (--numRounds2); - MM_OP_m (_mm_aesenc_si128, w[0]); - MM_OP_m (_mm_aesenclast_si128, w[1]); - MM_XOR (*data, m); - } - - p[-2] = ctr; -} - -#endif // USE_INTEL_VAES - -#else // USE_INTEL_AES - -/* no USE_INTEL_AES */ - -#pragma message("AES HW_SW stub was used") - -#define AES_TYPE_keys UInt32 -#define AES_TYPE_data Byte - -#define AES_FUNC_START(name) \ - void MY_FAST_CALL name(UInt32 *p, Byte *data, size_t numBlocks) \ - -#define AES_COMPAT_STUB(name) \ - AES_FUNC_START(name); \ - AES_FUNC_START(name ## _HW) \ - { name(p, data, numBlocks); } - -AES_COMPAT_STUB (AesCbc_Encode) -AES_COMPAT_STUB (AesCbc_Decode) -AES_COMPAT_STUB (AesCtr_Code) - -#endif // USE_INTEL_AES - - -#ifndef USE_INTEL_VAES - -#pragma message("VAES HW_SW stub was used") - -#define VAES_COMPAT_STUB(name) \ - void MY_FAST_CALL name ## _256(UInt32 *p, Byte *data, size_t numBlocks); \ - void MY_FAST_CALL name ## _256(UInt32 *p, Byte *data, size_t numBlocks) \ - { name((AES_TYPE_keys *)(void *)p, (AES_TYPE_data *)(void *)data, numBlocks); } - -VAES_COMPAT_STUB (AesCbc_Decode_HW) -VAES_COMPAT_STUB (AesCtr_Code_HW) - -#endif // ! USE_INTEL_VAES - - -#elif defined(MY_CPU_ARM_OR_ARM64) && defined(MY_CPU_LE) - - #if defined(__clang__) - #if (__clang_major__ >= 8) // fix that check - #define USE_HW_AES - #endif - #elif defined(__GNUC__) - #if (__GNUC__ >= 6) // fix that check - #define USE_HW_AES - #endif - #elif defined(_MSC_VER) - #if _MSC_VER >= 1910 - #define USE_HW_AES - #endif - #endif - -#ifdef USE_HW_AES - -// #pragma message("=== AES HW === ") - -#if defined(__clang__) || defined(__GNUC__) - #ifdef MY_CPU_ARM64 - #define ATTRIB_AES __attribute__((__target__("+crypto"))) - #else - #define ATTRIB_AES __attribute__((__target__("fpu=crypto-neon-fp-armv8"))) - #endif -#else - // _MSC_VER - // for arm32 - #define _ARM_USE_NEW_NEON_INTRINSICS -#endif - -#ifndef ATTRIB_AES - #define ATTRIB_AES -#endif - -#if defined(_MSC_VER) && defined(MY_CPU_ARM64) -#include -#else -#include -#endif - -typedef uint8x16_t v128; - -#define AES_FUNC_START(name) \ - void MY_FAST_CALL name(v128 *p, v128 *data, size_t numBlocks) - -#define AES_FUNC_START2(name) \ -AES_FUNC_START (name); \ -ATTRIB_AES \ -AES_FUNC_START (name) - -#define MM_OP(op, dest, src) dest = op(dest, src); -#define MM_OP_m(op, src) MM_OP(op, m, src); -#define MM_OP1_m(op) m = op(m); - -#define MM_XOR( dest, src) MM_OP(veorq_u8, dest, src); -#define MM_XOR_m( src) MM_XOR(m, src); - -#define AES_E_m(k) MM_OP_m (vaeseq_u8, k); -#define AES_E_MC_m(k) AES_E_m (k); MM_OP1_m(vaesmcq_u8); - - -AES_FUNC_START2 (AesCbc_Encode_HW) -{ - v128 m = *p; - const v128 k0 = p[2]; - const v128 k1 = p[3]; - const v128 k2 = p[4]; - const v128 k3 = p[5]; - const v128 k4 = p[6]; - const v128 k5 = p[7]; - const v128 k6 = p[8]; - const v128 k7 = p[9]; - const v128 k8 = p[10]; - const v128 k9 = p[11]; - const UInt32 numRounds2 = *(const UInt32 *)(p + 1); - const v128 *w = p + ((size_t)numRounds2 * 2); - const v128 k_z1 = w[1]; - const v128 k_z0 = w[2]; - for (; numBlocks != 0; numBlocks--, data++) - { - MM_XOR_m (*data); - AES_E_MC_m (k0) - AES_E_MC_m (k1) - AES_E_MC_m (k2) - AES_E_MC_m (k3) - AES_E_MC_m (k4) - AES_E_MC_m (k5) - AES_E_MC_m (k6) - AES_E_MC_m (k7) - AES_E_MC_m (k8) - if (numRounds2 >= 6) - { - AES_E_MC_m (k9) - AES_E_MC_m (p[12]) - if (numRounds2 != 6) - { - AES_E_MC_m (p[13]) - AES_E_MC_m (p[14]) - } - } - AES_E_m (k_z1); - MM_XOR_m (k_z0); - *data = m; - } - *p = m; -} - - -#define WOP_1(op) -#define WOP_2(op) WOP_1 (op) op (m1, 1); -#define WOP_3(op) WOP_2 (op) op (m2, 2); -#define WOP_4(op) WOP_3 (op) op (m3, 3); -#define WOP_5(op) WOP_4 (op) op (m4, 4); -#define WOP_6(op) WOP_5 (op) op (m5, 5); -#define WOP_7(op) WOP_6 (op) op (m6, 6); -#define WOP_8(op) WOP_7 (op) op (m7, 7); - - #define NUM_WAYS 8 - #define WOP_M1 WOP_8 - -#define WOP(op) op (m0, 0); WOP_M1(op) - -#define DECLARE_VAR(reg, ii) v128 reg -#define LOAD_data( reg, ii) reg = data[ii]; -#define STORE_data( reg, ii) data[ii] = reg; -#if (NUM_WAYS > 1) -#define XOR_data_M1(reg, ii) MM_XOR (reg, data[ii- 1]); -#endif - -#define MM_OP_key(op, reg) MM_OP (op, reg, key); - -#define AES_D_m(k) MM_OP_m (vaesdq_u8, k); -#define AES_D_IMC_m(k) AES_D_m (k); MM_OP1_m (vaesimcq_u8); - -#define AES_XOR( reg, ii) MM_OP_key (veorq_u8, reg) -#define AES_D( reg, ii) MM_OP_key (vaesdq_u8, reg) -#define AES_E( reg, ii) MM_OP_key (vaeseq_u8, reg) - -#define AES_D_IMC( reg, ii) AES_D (reg, ii); reg = vaesimcq_u8(reg) -#define AES_E_MC( reg, ii) AES_E (reg, ii); reg = vaesmcq_u8(reg) - -#define CTR_START(reg, ii) MM_OP (vaddq_u64, ctr, one); reg = vreinterpretq_u8_u64(ctr); -#define CTR_END( reg, ii) MM_XOR (data[ii], reg); - -#define WOP_KEY(op, n) { \ - const v128 key = w[n]; \ - WOP(op); } - -#define WIDE_LOOP_START \ - dataEnd = data + numBlocks; \ - if (numBlocks >= NUM_WAYS) \ - { dataEnd -= NUM_WAYS; do { \ - -#define WIDE_LOOP_END \ - data += NUM_WAYS; \ - } while (data <= dataEnd); \ - dataEnd += NUM_WAYS; } \ - -#define SINGLE_LOOP \ - for (; data < dataEnd; data++) - - -AES_FUNC_START2 (AesCbc_Decode_HW) -{ - v128 iv = *p; - const v128 *wStart = p + ((size_t)*(const UInt32 *)(p + 1)) * 2; - const v128 *dataEnd; - p += 2; - - WIDE_LOOP_START - { - const v128 *w = wStart; - WOP (DECLARE_VAR) - WOP (LOAD_data); - WOP_KEY (AES_D_IMC, 2) - do - { - WOP_KEY (AES_D_IMC, 1) - WOP_KEY (AES_D_IMC, 0) - w -= 2; - } - while (w != p); - WOP_KEY (AES_D, 1) - WOP_KEY (AES_XOR, 0) - MM_XOR (m0, iv); - WOP_M1 (XOR_data_M1) - iv = data[NUM_WAYS - 1]; - WOP (STORE_data); - } - WIDE_LOOP_END - - SINGLE_LOOP - { - const v128 *w = wStart; - v128 m = *data; - AES_D_IMC_m (w[2]) - do - { - AES_D_IMC_m (w[1]); - AES_D_IMC_m (w[0]); - w -= 2; - } - while (w != p); - AES_D_m (w[1]); - MM_XOR_m (w[0]); - MM_XOR_m (iv); - iv = *data; - *data = m; - } - - p[-2] = iv; -} - - -AES_FUNC_START2 (AesCtr_Code_HW) -{ - uint64x2_t ctr = vreinterpretq_u64_u8(*p); - const v128 *wEnd = p + ((size_t)*(const UInt32 *)(p + 1)) * 2; - const v128 *dataEnd; - uint64x2_t one = vdupq_n_u64(0); - one = vsetq_lane_u64(1, one, 0); - p += 2; - - WIDE_LOOP_START - { - const v128 *w = p; - WOP (DECLARE_VAR) - WOP (CTR_START); - do - { - WOP_KEY (AES_E_MC, 0) - WOP_KEY (AES_E_MC, 1) - w += 2; - } - while (w != wEnd); - WOP_KEY (AES_E_MC, 0) - WOP_KEY (AES_E, 1) - WOP_KEY (AES_XOR, 2) - WOP (CTR_END); - } - WIDE_LOOP_END - - SINGLE_LOOP - { - const v128 *w = p; - v128 m; - CTR_START (m, 0); - do - { - AES_E_MC_m (w[0]); - AES_E_MC_m (w[1]); - w += 2; - } - while (w != wEnd); - AES_E_MC_m (w[0]); - AES_E_m (w[1]); - MM_XOR_m (w[2]); - CTR_END (m, 0); - } - - p[-2] = vreinterpretq_u8_u64(ctr); -} - -#endif // USE_HW_AES - -#endif // MY_CPU_ARM_OR_ARM64 +/* AesOpt.c -- AES optimized code for x86 AES hardware instructions +2021-04-01 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" + +#ifdef MY_CPU_X86_OR_AMD64 + + #if defined(__clang__) + #if __clang_major__ > 3 || (__clang_major__ == 3 && __clang_minor__ >= 8) + #define USE_INTEL_AES + #define ATTRIB_AES __attribute__((__target__("aes"))) + #if (__clang_major__ >= 8) + #define USE_INTEL_VAES + #define ATTRIB_VAES __attribute__((__target__("aes,vaes,avx2"))) + #endif + #endif + #elif defined(__GNUC__) + #if (__GNUC__ > 4) || (__GNUC__ == 4 && __GNUC_MINOR__ >= 4) + #define USE_INTEL_AES + #ifndef __AES__ + #define ATTRIB_AES __attribute__((__target__("aes"))) + #endif + #if (__GNUC__ >= 8) + #define USE_INTEL_VAES + #define ATTRIB_VAES __attribute__((__target__("aes,vaes,avx2"))) + #endif + #endif + #elif defined(__INTEL_COMPILER) + #if (__INTEL_COMPILER >= 1110) + #define USE_INTEL_AES + #if (__INTEL_COMPILER >= 1900) + #define USE_INTEL_VAES + #endif + #endif + #elif defined(_MSC_VER) + #if (_MSC_VER > 1500) || (_MSC_FULL_VER >= 150030729) + #define USE_INTEL_AES + #if (_MSC_VER >= 1910) + #define USE_INTEL_VAES + #endif + #endif + #endif + +#ifndef ATTRIB_AES + #define ATTRIB_AES +#endif +#ifndef ATTRIB_VAES + #define ATTRIB_VAES +#endif + + +#ifdef USE_INTEL_AES + +#include + +#ifndef USE_INTEL_VAES +#define AES_TYPE_keys __m128i +#define AES_TYPE_data __m128i +#endif + +#define AES_FUNC_START(name) \ + void MY_FAST_CALL name(__m128i *p, __m128i *data, size_t numBlocks) + +#define AES_FUNC_START2(name) \ +AES_FUNC_START (name); \ +ATTRIB_AES \ +AES_FUNC_START (name) + +#define MM_OP(op, dest, src) dest = op(dest, src); +#define MM_OP_m(op, src) MM_OP(op, m, src); + +#define MM_XOR( dest, src) MM_OP(_mm_xor_si128, dest, src); +#define AVX_XOR(dest, src) MM_OP(_mm256_xor_si256, dest, src); + + +AES_FUNC_START2 (AesCbc_Encode_HW) +{ + __m128i m = *p; + const __m128i k0 = p[2]; + const __m128i k1 = p[3]; + const UInt32 numRounds2 = *(const UInt32 *)(p + 1) - 1; + for (; numBlocks != 0; numBlocks--, data++) + { + UInt32 r = numRounds2; + const __m128i *w = p + 4; + __m128i temp = *data; + MM_XOR (temp, k0); + MM_XOR (m, temp); + MM_OP_m (_mm_aesenc_si128, k1); + do + { + MM_OP_m (_mm_aesenc_si128, w[0]); + MM_OP_m (_mm_aesenc_si128, w[1]); + w += 2; + } + while (--r); + MM_OP_m (_mm_aesenclast_si128, w[0]); + *data = m; + } + *p = m; +} + + +#define WOP_1(op) +#define WOP_2(op) WOP_1 (op) op (m1, 1); +#define WOP_3(op) WOP_2 (op) op (m2, 2); +#define WOP_4(op) WOP_3 (op) op (m3, 3); +#ifdef MY_CPU_AMD64 +#define WOP_5(op) WOP_4 (op) op (m4, 4); +#define WOP_6(op) WOP_5 (op) op (m5, 5); +#define WOP_7(op) WOP_6 (op) op (m6, 6); +#define WOP_8(op) WOP_7 (op) op (m7, 7); +#endif +/* +#define WOP_9(op) WOP_8 (op) op (m8, 8); +#define WOP_10(op) WOP_9 (op) op (m9, 9); +#define WOP_11(op) WOP_10(op) op (m10, 10); +#define WOP_12(op) WOP_11(op) op (m11, 11); +#define WOP_13(op) WOP_12(op) op (m12, 12); +#define WOP_14(op) WOP_13(op) op (m13, 13); +*/ + +#ifdef MY_CPU_AMD64 + #define NUM_WAYS 8 + #define WOP_M1 WOP_8 +#else + #define NUM_WAYS 4 + #define WOP_M1 WOP_4 +#endif + +#define WOP(op) op (m0, 0); WOP_M1(op) + + +#define DECLARE_VAR(reg, ii) __m128i reg +#define LOAD_data( reg, ii) reg = data[ii]; +#define STORE_data( reg, ii) data[ii] = reg; +#if (NUM_WAYS > 1) +#define XOR_data_M1(reg, ii) MM_XOR (reg, data[ii- 1]); +#endif + +#define AVX__DECLARE_VAR(reg, ii) __m256i reg +#define AVX__LOAD_data( reg, ii) reg = ((const __m256i *)(const void *)data)[ii]; +#define AVX__STORE_data( reg, ii) ((__m256i *)(void *)data)[ii] = reg; +#define AVX__XOR_data_M1(reg, ii) AVX_XOR (reg, (((const __m256i *)(const void *)(data - 1))[ii])); + +#define MM_OP_key(op, reg) MM_OP(op, reg, key); + +#define AES_DEC( reg, ii) MM_OP_key (_mm_aesdec_si128, reg) +#define AES_DEC_LAST( reg, ii) MM_OP_key (_mm_aesdeclast_si128, reg) +#define AES_ENC( reg, ii) MM_OP_key (_mm_aesenc_si128, reg) +#define AES_ENC_LAST( reg, ii) MM_OP_key (_mm_aesenclast_si128, reg) +#define AES_XOR( reg, ii) MM_OP_key (_mm_xor_si128, reg) + + +#define AVX__AES_DEC( reg, ii) MM_OP_key (_mm256_aesdec_epi128, reg) +#define AVX__AES_DEC_LAST( reg, ii) MM_OP_key (_mm256_aesdeclast_epi128, reg) +#define AVX__AES_ENC( reg, ii) MM_OP_key (_mm256_aesenc_epi128, reg) +#define AVX__AES_ENC_LAST( reg, ii) MM_OP_key (_mm256_aesenclast_epi128, reg) +#define AVX__AES_XOR( reg, ii) MM_OP_key (_mm256_xor_si256, reg) + +#define CTR_START(reg, ii) MM_OP (_mm_add_epi64, ctr, one); reg = ctr; +#define CTR_END( reg, ii) MM_XOR (data[ii], reg); + +#define AVX__CTR_START(reg, ii) MM_OP (_mm256_add_epi64, ctr2, two); reg = _mm256_xor_si256(ctr2, key); +#define AVX__CTR_END( reg, ii) AVX_XOR (((__m256i *)(void *)data)[ii], reg); + +#define WOP_KEY(op, n) { \ + const __m128i key = w[n]; \ + WOP(op); } + +#define AVX__WOP_KEY(op, n) { \ + const __m256i key = w[n]; \ + WOP(op); } + + +#define WIDE_LOOP_START \ + dataEnd = data + numBlocks; \ + if (numBlocks >= NUM_WAYS) \ + { dataEnd -= NUM_WAYS; do { \ + + +#define WIDE_LOOP_END \ + data += NUM_WAYS; \ + } while (data <= dataEnd); \ + dataEnd += NUM_WAYS; } \ + + +#define SINGLE_LOOP \ + for (; data < dataEnd; data++) + + +#define NUM_AES_KEYS_MAX 15 + +#define WIDE_LOOP_START_AVX(OP) \ + dataEnd = data + numBlocks; \ + if (numBlocks >= NUM_WAYS * 2) \ + { __m256i keys[NUM_AES_KEYS_MAX]; \ + UInt32 ii; \ + OP \ + for (ii = 0; ii < numRounds; ii++) \ + keys[ii] = _mm256_broadcastsi128_si256(p[ii]); \ + dataEnd -= NUM_WAYS * 2; do { \ + + +#define WIDE_LOOP_END_AVX(OP) \ + data += NUM_WAYS * 2; \ + } while (data <= dataEnd); \ + dataEnd += NUM_WAYS * 2; \ + OP \ + _mm256_zeroupper(); \ + } \ + +/* MSVC for x86: If we don't call _mm256_zeroupper(), and -arch:IA32 is not specified, + MSVC still can insert vzeroupper instruction. */ + + +AES_FUNC_START2 (AesCbc_Decode_HW) +{ + __m128i iv = *p; + const __m128i *wStart = p + *(const UInt32 *)(p + 1) * 2 + 2 - 1; + const __m128i *dataEnd; + p += 2; + + WIDE_LOOP_START + { + const __m128i *w = wStart; + + WOP (DECLARE_VAR) + WOP (LOAD_data); + WOP_KEY (AES_XOR, 1) + + do + { + WOP_KEY (AES_DEC, 0) + w--; + } + while (w != p); + WOP_KEY (AES_DEC_LAST, 0) + + MM_XOR (m0, iv); + WOP_M1 (XOR_data_M1) + iv = data[NUM_WAYS - 1]; + WOP (STORE_data); + } + WIDE_LOOP_END + + SINGLE_LOOP + { + const __m128i *w = wStart - 1; + __m128i m = _mm_xor_si128 (w[2], *data); + do + { + MM_OP_m (_mm_aesdec_si128, w[1]); + MM_OP_m (_mm_aesdec_si128, w[0]); + w -= 2; + } + while (w != p); + MM_OP_m (_mm_aesdec_si128, w[1]); + MM_OP_m (_mm_aesdeclast_si128, w[0]); + + MM_XOR (m, iv); + iv = *data; + *data = m; + } + + p[-2] = iv; +} + + +AES_FUNC_START2 (AesCtr_Code_HW) +{ + __m128i ctr = *p; + UInt32 numRoundsMinus2 = *(const UInt32 *)(p + 1) * 2 - 1; + const __m128i *dataEnd; + __m128i one = _mm_cvtsi32_si128(1); + + p += 2; + + WIDE_LOOP_START + { + const __m128i *w = p; + UInt32 r = numRoundsMinus2; + WOP (DECLARE_VAR) + WOP (CTR_START); + WOP_KEY (AES_XOR, 0) + w += 1; + do + { + WOP_KEY (AES_ENC, 0) + w += 1; + } + while (--r); + WOP_KEY (AES_ENC_LAST, 0) + + WOP (CTR_END); + } + WIDE_LOOP_END + + SINGLE_LOOP + { + UInt32 numRounds2 = *(const UInt32 *)(p - 2 + 1) - 1; + const __m128i *w = p; + __m128i m; + MM_OP (_mm_add_epi64, ctr, one); + m = _mm_xor_si128 (ctr, p[0]); + w += 1; + do + { + MM_OP_m (_mm_aesenc_si128, w[0]); + MM_OP_m (_mm_aesenc_si128, w[1]); + w += 2; + } + while (--numRounds2); + MM_OP_m (_mm_aesenc_si128, w[0]); + MM_OP_m (_mm_aesenclast_si128, w[1]); + MM_XOR (*data, m); + } + + p[-2] = ctr; +} + + + +#ifdef USE_INTEL_VAES + +#if defined(__clang__) && defined(_MSC_VER) +#define __SSE4_2__ +#define __AES__ +#define __AVX__ +#define __AVX2__ +#define __VAES__ +#define __AVX512F__ +#define __AVX512VL__ +#endif + +#include + +#define VAES_FUNC_START2(name) \ +AES_FUNC_START (name); \ +ATTRIB_VAES \ +AES_FUNC_START (name) + +VAES_FUNC_START2 (AesCbc_Decode_HW_256) +{ + __m128i iv = *p; + const __m128i *dataEnd; + UInt32 numRounds = *(const UInt32 *)(p + 1) * 2 + 1; + p += 2; + + WIDE_LOOP_START_AVX(;) + { + const __m256i *w = keys + numRounds - 2; + + WOP (AVX__DECLARE_VAR) + WOP (AVX__LOAD_data); + AVX__WOP_KEY (AVX__AES_XOR, 1) + + do + { + AVX__WOP_KEY (AVX__AES_DEC, 0) + w--; + } + while (w != keys); + AVX__WOP_KEY (AVX__AES_DEC_LAST, 0) + + AVX_XOR (m0, _mm256_setr_m128i(iv, data[0])); + WOP_M1 (AVX__XOR_data_M1) + iv = data[NUM_WAYS * 2 - 1]; + WOP (AVX__STORE_data); + } + WIDE_LOOP_END_AVX(;) + + SINGLE_LOOP + { + const __m128i *w = p + *(const UInt32 *)(p + 1 - 2) * 2 + 1 - 3; + __m128i m = _mm_xor_si128 (w[2], *data); + do + { + MM_OP_m (_mm_aesdec_si128, w[1]); + MM_OP_m (_mm_aesdec_si128, w[0]); + w -= 2; + } + while (w != p); + MM_OP_m (_mm_aesdec_si128, w[1]); + MM_OP_m (_mm_aesdeclast_si128, w[0]); + + MM_XOR (m, iv); + iv = *data; + *data = m; + } + + p[-2] = iv; +} + + +/* +SSE2: _mm_cvtsi32_si128 : movd +AVX: _mm256_setr_m128i : vinsertf128 +AVX2: _mm256_add_epi64 : vpaddq ymm, ymm, ymm + _mm256_extracti128_si256 : vextracti128 + _mm256_broadcastsi128_si256 : vbroadcasti128 +*/ + +#define AVX__CTR_LOOP_START \ + ctr2 = _mm256_setr_m128i(_mm_sub_epi64(ctr, one), ctr); \ + two = _mm256_setr_m128i(one, one); \ + two = _mm256_add_epi64(two, two); \ + +// two = _mm256_setr_epi64x(2, 0, 2, 0); + +#define AVX__CTR_LOOP_ENC \ + ctr = _mm256_extracti128_si256 (ctr2, 1); \ + +VAES_FUNC_START2 (AesCtr_Code_HW_256) +{ + __m128i ctr = *p; + UInt32 numRounds = *(const UInt32 *)(p + 1) * 2 + 1; + const __m128i *dataEnd; + __m128i one = _mm_cvtsi32_si128(1); + __m256i ctr2, two; + p += 2; + + WIDE_LOOP_START_AVX (AVX__CTR_LOOP_START) + { + const __m256i *w = keys; + UInt32 r = numRounds - 2; + WOP (AVX__DECLARE_VAR) + AVX__WOP_KEY (AVX__CTR_START, 0); + + w += 1; + do + { + AVX__WOP_KEY (AVX__AES_ENC, 0) + w += 1; + } + while (--r); + AVX__WOP_KEY (AVX__AES_ENC_LAST, 0) + + WOP (AVX__CTR_END); + } + WIDE_LOOP_END_AVX (AVX__CTR_LOOP_ENC) + + SINGLE_LOOP + { + UInt32 numRounds2 = *(const UInt32 *)(p - 2 + 1) - 1; + const __m128i *w = p; + __m128i m; + MM_OP (_mm_add_epi64, ctr, one); + m = _mm_xor_si128 (ctr, p[0]); + w += 1; + do + { + MM_OP_m (_mm_aesenc_si128, w[0]); + MM_OP_m (_mm_aesenc_si128, w[1]); + w += 2; + } + while (--numRounds2); + MM_OP_m (_mm_aesenc_si128, w[0]); + MM_OP_m (_mm_aesenclast_si128, w[1]); + MM_XOR (*data, m); + } + + p[-2] = ctr; +} + +#endif // USE_INTEL_VAES + +#else // USE_INTEL_AES + +/* no USE_INTEL_AES */ + +#pragma message("AES HW_SW stub was used") + +#define AES_TYPE_keys UInt32 +#define AES_TYPE_data Byte + +#define AES_FUNC_START(name) \ + void MY_FAST_CALL name(UInt32 *p, Byte *data, size_t numBlocks) \ + +#define AES_COMPAT_STUB(name) \ + AES_FUNC_START(name); \ + AES_FUNC_START(name ## _HW) \ + { name(p, data, numBlocks); } + +AES_COMPAT_STUB (AesCbc_Encode) +AES_COMPAT_STUB (AesCbc_Decode) +AES_COMPAT_STUB (AesCtr_Code) + +#endif // USE_INTEL_AES + + +#ifndef USE_INTEL_VAES + +#pragma message("VAES HW_SW stub was used") + +#define VAES_COMPAT_STUB(name) \ + void MY_FAST_CALL name ## _256(UInt32 *p, Byte *data, size_t numBlocks); \ + void MY_FAST_CALL name ## _256(UInt32 *p, Byte *data, size_t numBlocks) \ + { name((AES_TYPE_keys *)(void *)p, (AES_TYPE_data *)(void *)data, numBlocks); } + +VAES_COMPAT_STUB (AesCbc_Decode_HW) +VAES_COMPAT_STUB (AesCtr_Code_HW) + +#endif // ! USE_INTEL_VAES + + +#elif defined(MY_CPU_ARM_OR_ARM64) && defined(MY_CPU_LE) + + #if defined(__clang__) + #if (__clang_major__ >= 8) // fix that check + #define USE_HW_AES + #endif + #elif defined(__GNUC__) + #if (__GNUC__ >= 6) // fix that check + #define USE_HW_AES + #endif + #elif defined(_MSC_VER) + #if _MSC_VER >= 1910 + #define USE_HW_AES + #endif + #endif + +#ifdef USE_HW_AES + +// #pragma message("=== AES HW === ") + +#if defined(__clang__) || defined(__GNUC__) + #ifdef MY_CPU_ARM64 + #define ATTRIB_AES __attribute__((__target__("+crypto"))) + #else + #define ATTRIB_AES __attribute__((__target__("fpu=crypto-neon-fp-armv8"))) + #endif +#else + // _MSC_VER + // for arm32 + #define _ARM_USE_NEW_NEON_INTRINSICS +#endif + +#ifndef ATTRIB_AES + #define ATTRIB_AES +#endif + +#if defined(_MSC_VER) && defined(MY_CPU_ARM64) +#include +#else +#include +#endif + +typedef uint8x16_t v128; + +#define AES_FUNC_START(name) \ + void MY_FAST_CALL name(v128 *p, v128 *data, size_t numBlocks) + +#define AES_FUNC_START2(name) \ +AES_FUNC_START (name); \ +ATTRIB_AES \ +AES_FUNC_START (name) + +#define MM_OP(op, dest, src) dest = op(dest, src); +#define MM_OP_m(op, src) MM_OP(op, m, src); +#define MM_OP1_m(op) m = op(m); + +#define MM_XOR( dest, src) MM_OP(veorq_u8, dest, src); +#define MM_XOR_m( src) MM_XOR(m, src); + +#define AES_E_m(k) MM_OP_m (vaeseq_u8, k); +#define AES_E_MC_m(k) AES_E_m (k); MM_OP1_m(vaesmcq_u8); + + +AES_FUNC_START2 (AesCbc_Encode_HW) +{ + v128 m = *p; + const v128 k0 = p[2]; + const v128 k1 = p[3]; + const v128 k2 = p[4]; + const v128 k3 = p[5]; + const v128 k4 = p[6]; + const v128 k5 = p[7]; + const v128 k6 = p[8]; + const v128 k7 = p[9]; + const v128 k8 = p[10]; + const v128 k9 = p[11]; + const UInt32 numRounds2 = *(const UInt32 *)(p + 1); + const v128 *w = p + ((size_t)numRounds2 * 2); + const v128 k_z1 = w[1]; + const v128 k_z0 = w[2]; + for (; numBlocks != 0; numBlocks--, data++) + { + MM_XOR_m (*data); + AES_E_MC_m (k0) + AES_E_MC_m (k1) + AES_E_MC_m (k2) + AES_E_MC_m (k3) + AES_E_MC_m (k4) + AES_E_MC_m (k5) + AES_E_MC_m (k6) + AES_E_MC_m (k7) + AES_E_MC_m (k8) + if (numRounds2 >= 6) + { + AES_E_MC_m (k9) + AES_E_MC_m (p[12]) + if (numRounds2 != 6) + { + AES_E_MC_m (p[13]) + AES_E_MC_m (p[14]) + } + } + AES_E_m (k_z1); + MM_XOR_m (k_z0); + *data = m; + } + *p = m; +} + + +#define WOP_1(op) +#define WOP_2(op) WOP_1 (op) op (m1, 1); +#define WOP_3(op) WOP_2 (op) op (m2, 2); +#define WOP_4(op) WOP_3 (op) op (m3, 3); +#define WOP_5(op) WOP_4 (op) op (m4, 4); +#define WOP_6(op) WOP_5 (op) op (m5, 5); +#define WOP_7(op) WOP_6 (op) op (m6, 6); +#define WOP_8(op) WOP_7 (op) op (m7, 7); + + #define NUM_WAYS 8 + #define WOP_M1 WOP_8 + +#define WOP(op) op (m0, 0); WOP_M1(op) + +#define DECLARE_VAR(reg, ii) v128 reg +#define LOAD_data( reg, ii) reg = data[ii]; +#define STORE_data( reg, ii) data[ii] = reg; +#if (NUM_WAYS > 1) +#define XOR_data_M1(reg, ii) MM_XOR (reg, data[ii- 1]); +#endif + +#define MM_OP_key(op, reg) MM_OP (op, reg, key); + +#define AES_D_m(k) MM_OP_m (vaesdq_u8, k); +#define AES_D_IMC_m(k) AES_D_m (k); MM_OP1_m (vaesimcq_u8); + +#define AES_XOR( reg, ii) MM_OP_key (veorq_u8, reg) +#define AES_D( reg, ii) MM_OP_key (vaesdq_u8, reg) +#define AES_E( reg, ii) MM_OP_key (vaeseq_u8, reg) + +#define AES_D_IMC( reg, ii) AES_D (reg, ii); reg = vaesimcq_u8(reg) +#define AES_E_MC( reg, ii) AES_E (reg, ii); reg = vaesmcq_u8(reg) + +#define CTR_START(reg, ii) MM_OP (vaddq_u64, ctr, one); reg = vreinterpretq_u8_u64(ctr); +#define CTR_END( reg, ii) MM_XOR (data[ii], reg); + +#define WOP_KEY(op, n) { \ + const v128 key = w[n]; \ + WOP(op); } + +#define WIDE_LOOP_START \ + dataEnd = data + numBlocks; \ + if (numBlocks >= NUM_WAYS) \ + { dataEnd -= NUM_WAYS; do { \ + +#define WIDE_LOOP_END \ + data += NUM_WAYS; \ + } while (data <= dataEnd); \ + dataEnd += NUM_WAYS; } \ + +#define SINGLE_LOOP \ + for (; data < dataEnd; data++) + + +AES_FUNC_START2 (AesCbc_Decode_HW) +{ + v128 iv = *p; + const v128 *wStart = p + ((size_t)*(const UInt32 *)(p + 1)) * 2; + const v128 *dataEnd; + p += 2; + + WIDE_LOOP_START + { + const v128 *w = wStart; + WOP (DECLARE_VAR) + WOP (LOAD_data); + WOP_KEY (AES_D_IMC, 2) + do + { + WOP_KEY (AES_D_IMC, 1) + WOP_KEY (AES_D_IMC, 0) + w -= 2; + } + while (w != p); + WOP_KEY (AES_D, 1) + WOP_KEY (AES_XOR, 0) + MM_XOR (m0, iv); + WOP_M1 (XOR_data_M1) + iv = data[NUM_WAYS - 1]; + WOP (STORE_data); + } + WIDE_LOOP_END + + SINGLE_LOOP + { + const v128 *w = wStart; + v128 m = *data; + AES_D_IMC_m (w[2]) + do + { + AES_D_IMC_m (w[1]); + AES_D_IMC_m (w[0]); + w -= 2; + } + while (w != p); + AES_D_m (w[1]); + MM_XOR_m (w[0]); + MM_XOR_m (iv); + iv = *data; + *data = m; + } + + p[-2] = iv; +} + + +AES_FUNC_START2 (AesCtr_Code_HW) +{ + uint64x2_t ctr = vreinterpretq_u64_u8(*p); + const v128 *wEnd = p + ((size_t)*(const UInt32 *)(p + 1)) * 2; + const v128 *dataEnd; + uint64x2_t one = vdupq_n_u64(0); + one = vsetq_lane_u64(1, one, 0); + p += 2; + + WIDE_LOOP_START + { + const v128 *w = p; + WOP (DECLARE_VAR) + WOP (CTR_START); + do + { + WOP_KEY (AES_E_MC, 0) + WOP_KEY (AES_E_MC, 1) + w += 2; + } + while (w != wEnd); + WOP_KEY (AES_E_MC, 0) + WOP_KEY (AES_E, 1) + WOP_KEY (AES_XOR, 2) + WOP (CTR_END); + } + WIDE_LOOP_END + + SINGLE_LOOP + { + const v128 *w = p; + v128 m; + CTR_START (m, 0); + do + { + AES_E_MC_m (w[0]); + AES_E_MC_m (w[1]); + w += 2; + } + while (w != wEnd); + AES_E_MC_m (w[0]); + AES_E_m (w[1]); + MM_XOR_m (w[2]); + CTR_END (m, 0); + } + + p[-2] = vreinterpretq_u8_u64(ctr); +} + +#endif // USE_HW_AES + +#endif // MY_CPU_ARM_OR_ARM64 diff --git a/deps/LZMA-SDK/C/Alloc.c b/deps/LZMA-SDK/C/Alloc.c index 0d0107c56..242ae3d4f 100644 --- a/deps/LZMA-SDK/C/Alloc.c +++ b/deps/LZMA-SDK/C/Alloc.c @@ -1,463 +1,463 @@ -/* Alloc.c -- Memory allocation functions -2021-07-13 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include - -#ifdef _WIN32 -#include -#endif -#include - -#include "Alloc.h" - -/* #define _SZ_ALLOC_DEBUG */ - -/* use _SZ_ALLOC_DEBUG to debug alloc/free operations */ -#ifdef _SZ_ALLOC_DEBUG - -#include -int g_allocCount = 0; -int g_allocCountMid = 0; -int g_allocCountBig = 0; - - -#define CONVERT_INT_TO_STR(charType, tempSize) \ - unsigned char temp[tempSize]; unsigned i = 0; \ - while (val >= 10) { temp[i++] = (unsigned char)('0' + (unsigned)(val % 10)); val /= 10; } \ - *s++ = (charType)('0' + (unsigned)val); \ - while (i != 0) { i--; *s++ = temp[i]; } \ - *s = 0; - -static void ConvertUInt64ToString(UInt64 val, char *s) -{ - CONVERT_INT_TO_STR(char, 24); -} - -#define GET_HEX_CHAR(t) ((char)(((t < 10) ? ('0' + t) : ('A' + (t - 10))))) - -static void ConvertUInt64ToHex(UInt64 val, char *s) -{ - UInt64 v = val; - unsigned i; - for (i = 1;; i++) - { - v >>= 4; - if (v == 0) - break; - } - s[i] = 0; - do - { - unsigned t = (unsigned)(val & 0xF); - val >>= 4; - s[--i] = GET_HEX_CHAR(t); - } - while (i); -} - -#define DEBUG_OUT_STREAM stderr - -static void Print(const char *s) -{ - fputs(s, DEBUG_OUT_STREAM); -} - -static void PrintAligned(const char *s, size_t align) -{ - size_t len = strlen(s); - for(;;) - { - fputc(' ', DEBUG_OUT_STREAM); - if (len >= align) - break; - ++len; - } - Print(s); -} - -static void PrintLn() -{ - Print("\n"); -} - -static void PrintHex(UInt64 v, size_t align) -{ - char s[32]; - ConvertUInt64ToHex(v, s); - PrintAligned(s, align); -} - -static void PrintDec(UInt64 v, size_t align) -{ - char s[32]; - ConvertUInt64ToString(v, s); - PrintAligned(s, align); -} - -static void PrintAddr(void *p) -{ - PrintHex((UInt64)(size_t)(ptrdiff_t)p, 12); -} - - -#define PRINT_ALLOC(name, cnt, size, ptr) \ - Print(name " "); \ - PrintDec(cnt++, 10); \ - PrintHex(size, 10); \ - PrintAddr(ptr); \ - PrintLn(); - -#define PRINT_FREE(name, cnt, ptr) if (ptr) { \ - Print(name " "); \ - PrintDec(--cnt, 10); \ - PrintAddr(ptr); \ - PrintLn(); } - -#else - -#define PRINT_ALLOC(name, cnt, size, ptr) -#define PRINT_FREE(name, cnt, ptr) -#define Print(s) -#define PrintLn() -#define PrintHex(v, align) -#define PrintAddr(p) - -#endif - - - -void *MyAlloc(size_t size) -{ - if (size == 0) - return NULL; - PRINT_ALLOC("Alloc ", g_allocCount, size, NULL); - #ifdef _SZ_ALLOC_DEBUG - { - void *p = malloc(size); - // PRINT_ALLOC("Alloc ", g_allocCount, size, p); - return p; - } - #else - return malloc(size); - #endif -} - -void MyFree(void *address) -{ - PRINT_FREE("Free ", g_allocCount, address); - - free(address); -} - -#ifdef _WIN32 - -void *MidAlloc(size_t size) -{ - if (size == 0) - return NULL; - - PRINT_ALLOC("Alloc-Mid", g_allocCountMid, size, NULL); - - return VirtualAlloc(NULL, size, MEM_COMMIT, PAGE_READWRITE); -} - -void MidFree(void *address) -{ - PRINT_FREE("Free-Mid", g_allocCountMid, address); - - if (!address) - return; - VirtualFree(address, 0, MEM_RELEASE); -} - -#ifdef _7ZIP_LARGE_PAGES - -#ifdef MEM_LARGE_PAGES - #define MY__MEM_LARGE_PAGES MEM_LARGE_PAGES -#else - #define MY__MEM_LARGE_PAGES 0x20000000 -#endif - -extern -SIZE_T g_LargePageSize; -SIZE_T g_LargePageSize = 0; -typedef SIZE_T (WINAPI *GetLargePageMinimumP)(VOID); - -#endif // _7ZIP_LARGE_PAGES - -void SetLargePageSize() -{ - #ifdef _7ZIP_LARGE_PAGES - SIZE_T size; - GetLargePageMinimumP largePageMinimum = (GetLargePageMinimumP) - GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "GetLargePageMinimum"); - if (!largePageMinimum) - return; - size = largePageMinimum(); - if (size == 0 || (size & (size - 1)) != 0) - return; - g_LargePageSize = size; - #endif -} - - -void *BigAlloc(size_t size) -{ - if (size == 0) - return NULL; - - PRINT_ALLOC("Alloc-Big", g_allocCountBig, size, NULL); - - #ifdef _7ZIP_LARGE_PAGES - { - SIZE_T ps = g_LargePageSize; - if (ps != 0 && ps <= (1 << 30) && size > (ps / 2)) - { - size_t size2; - ps--; - size2 = (size + ps) & ~ps; - if (size2 >= size) - { - void *res = VirtualAlloc(NULL, size2, MEM_COMMIT | MY__MEM_LARGE_PAGES, PAGE_READWRITE); - if (res) - return res; - } - } - } - #endif - - return VirtualAlloc(NULL, size, MEM_COMMIT, PAGE_READWRITE); -} - -void BigFree(void *address) -{ - PRINT_FREE("Free-Big", g_allocCountBig, address); - - if (!address) - return; - VirtualFree(address, 0, MEM_RELEASE); -} - -#endif - - -static void *SzAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return MyAlloc(size); } -static void SzFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); MyFree(address); } -const ISzAlloc g_Alloc = { SzAlloc, SzFree }; - -#ifdef _WIN32 -static void *SzMidAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return MidAlloc(size); } -static void SzMidFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); MidFree(address); } -static void *SzBigAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return BigAlloc(size); } -static void SzBigFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); BigFree(address); } -const ISzAlloc g_MidAlloc = { SzMidAlloc, SzMidFree }; -const ISzAlloc g_BigAlloc = { SzBigAlloc, SzBigFree }; -#endif - -/* - uintptr_t : C99 (optional) - : unsupported in VS6 -*/ - -#ifdef _WIN32 - typedef UINT_PTR UIntPtr; -#else - /* - typedef uintptr_t UIntPtr; - */ - typedef ptrdiff_t UIntPtr; -#endif - - -#define ADJUST_ALLOC_SIZE 0 -/* -#define ADJUST_ALLOC_SIZE (sizeof(void *) - 1) -*/ -/* - Use (ADJUST_ALLOC_SIZE = (sizeof(void *) - 1)), if - MyAlloc() can return address that is NOT multiple of sizeof(void *). -*/ - - -/* -#define MY_ALIGN_PTR_DOWN(p, align) ((void *)((char *)(p) - ((size_t)(UIntPtr)(p) & ((align) - 1)))) -*/ -#define MY_ALIGN_PTR_DOWN(p, align) ((void *)((((UIntPtr)(p)) & ~((UIntPtr)(align) - 1)))) - - -#if !defined(_WIN32) && defined(_POSIX_C_SOURCE) && (_POSIX_C_SOURCE >= 200112L) - #define USE_posix_memalign -#endif - -#ifndef USE_posix_memalign -#define MY_ALIGN_PTR_UP_PLUS(p, align) MY_ALIGN_PTR_DOWN(((char *)(p) + (align) + ADJUST_ALLOC_SIZE), align) -#endif - -/* - This posix_memalign() is for test purposes only. - We also need special Free() function instead of free(), - if this posix_memalign() is used. -*/ - -/* -static int posix_memalign(void **ptr, size_t align, size_t size) -{ - size_t newSize = size + align; - void *p; - void *pAligned; - *ptr = NULL; - if (newSize < size) - return 12; // ENOMEM - p = MyAlloc(newSize); - if (!p) - return 12; // ENOMEM - pAligned = MY_ALIGN_PTR_UP_PLUS(p, align); - ((void **)pAligned)[-1] = p; - *ptr = pAligned; - return 0; -} -*/ - -/* - ALLOC_ALIGN_SIZE >= sizeof(void *) - ALLOC_ALIGN_SIZE >= cache_line_size -*/ - -#define ALLOC_ALIGN_SIZE ((size_t)1 << 7) - -static void *SzAlignedAlloc(ISzAllocPtr pp, size_t size) -{ - #ifndef USE_posix_memalign - - void *p; - void *pAligned; - size_t newSize; - UNUSED_VAR(pp); - - /* also we can allocate additional dummy ALLOC_ALIGN_SIZE bytes after aligned - block to prevent cache line sharing with another allocated blocks */ - - newSize = size + ALLOC_ALIGN_SIZE * 1 + ADJUST_ALLOC_SIZE; - if (newSize < size) - return NULL; - - p = MyAlloc(newSize); - - if (!p) - return NULL; - pAligned = MY_ALIGN_PTR_UP_PLUS(p, ALLOC_ALIGN_SIZE); - - Print(" size="); PrintHex(size, 8); - Print(" a_size="); PrintHex(newSize, 8); - Print(" ptr="); PrintAddr(p); - Print(" a_ptr="); PrintAddr(pAligned); - PrintLn(); - - ((void **)pAligned)[-1] = p; - - return pAligned; - - #else - - void *p; - UNUSED_VAR(pp); - if (posix_memalign(&p, ALLOC_ALIGN_SIZE, size)) - return NULL; - - Print(" posix_memalign="); PrintAddr(p); - PrintLn(); - - return p; - - #endif -} - - -static void SzAlignedFree(ISzAllocPtr pp, void *address) -{ - UNUSED_VAR(pp); - #ifndef USE_posix_memalign - if (address) - MyFree(((void **)address)[-1]); - #else - free(address); - #endif -} - - -const ISzAlloc g_AlignedAlloc = { SzAlignedAlloc, SzAlignedFree }; - - - -#define MY_ALIGN_PTR_DOWN_1(p) MY_ALIGN_PTR_DOWN(p, sizeof(void *)) - -/* we align ptr to support cases where CAlignOffsetAlloc::offset is not multiply of sizeof(void *) */ -#define REAL_BLOCK_PTR_VAR(p) ((void **)MY_ALIGN_PTR_DOWN_1(p))[-1] -/* -#define REAL_BLOCK_PTR_VAR(p) ((void **)(p))[-1] -*/ - -static void *AlignOffsetAlloc_Alloc(ISzAllocPtr pp, size_t size) -{ - CAlignOffsetAlloc *p = CONTAINER_FROM_VTBL(pp, CAlignOffsetAlloc, vt); - void *adr; - void *pAligned; - size_t newSize; - size_t extra; - size_t alignSize = (size_t)1 << p->numAlignBits; - - if (alignSize < sizeof(void *)) - alignSize = sizeof(void *); - - if (p->offset >= alignSize) - return NULL; - - /* also we can allocate additional dummy ALLOC_ALIGN_SIZE bytes after aligned - block to prevent cache line sharing with another allocated blocks */ - extra = p->offset & (sizeof(void *) - 1); - newSize = size + alignSize + extra + ADJUST_ALLOC_SIZE; - if (newSize < size) - return NULL; - - adr = ISzAlloc_Alloc(p->baseAlloc, newSize); - - if (!adr) - return NULL; - - pAligned = (char *)MY_ALIGN_PTR_DOWN((char *)adr + - alignSize - p->offset + extra + ADJUST_ALLOC_SIZE, alignSize) + p->offset; - - PrintLn(); - Print("- Aligned: "); - Print(" size="); PrintHex(size, 8); - Print(" a_size="); PrintHex(newSize, 8); - Print(" ptr="); PrintAddr(adr); - Print(" a_ptr="); PrintAddr(pAligned); - PrintLn(); - - REAL_BLOCK_PTR_VAR(pAligned) = adr; - - return pAligned; -} - - -static void AlignOffsetAlloc_Free(ISzAllocPtr pp, void *address) -{ - if (address) - { - CAlignOffsetAlloc *p = CONTAINER_FROM_VTBL(pp, CAlignOffsetAlloc, vt); - PrintLn(); - Print("- Aligned Free: "); - PrintLn(); - ISzAlloc_Free(p->baseAlloc, REAL_BLOCK_PTR_VAR(address)); - } -} - - -void AlignOffsetAlloc_CreateVTable(CAlignOffsetAlloc *p) -{ - p->vt.Alloc = AlignOffsetAlloc_Alloc; - p->vt.Free = AlignOffsetAlloc_Free; -} +/* Alloc.c -- Memory allocation functions +2021-07-13 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#ifdef _WIN32 +#include +#endif +#include + +#include "Alloc.h" + +/* #define _SZ_ALLOC_DEBUG */ + +/* use _SZ_ALLOC_DEBUG to debug alloc/free operations */ +#ifdef _SZ_ALLOC_DEBUG + +#include +int g_allocCount = 0; +int g_allocCountMid = 0; +int g_allocCountBig = 0; + + +#define CONVERT_INT_TO_STR(charType, tempSize) \ + unsigned char temp[tempSize]; unsigned i = 0; \ + while (val >= 10) { temp[i++] = (unsigned char)('0' + (unsigned)(val % 10)); val /= 10; } \ + *s++ = (charType)('0' + (unsigned)val); \ + while (i != 0) { i--; *s++ = temp[i]; } \ + *s = 0; + +static void ConvertUInt64ToString(UInt64 val, char *s) +{ + CONVERT_INT_TO_STR(char, 24); +} + +#define GET_HEX_CHAR(t) ((char)(((t < 10) ? ('0' + t) : ('A' + (t - 10))))) + +static void ConvertUInt64ToHex(UInt64 val, char *s) +{ + UInt64 v = val; + unsigned i; + for (i = 1;; i++) + { + v >>= 4; + if (v == 0) + break; + } + s[i] = 0; + do + { + unsigned t = (unsigned)(val & 0xF); + val >>= 4; + s[--i] = GET_HEX_CHAR(t); + } + while (i); +} + +#define DEBUG_OUT_STREAM stderr + +static void Print(const char *s) +{ + fputs(s, DEBUG_OUT_STREAM); +} + +static void PrintAligned(const char *s, size_t align) +{ + size_t len = strlen(s); + for(;;) + { + fputc(' ', DEBUG_OUT_STREAM); + if (len >= align) + break; + ++len; + } + Print(s); +} + +static void PrintLn() +{ + Print("\n"); +} + +static void PrintHex(UInt64 v, size_t align) +{ + char s[32]; + ConvertUInt64ToHex(v, s); + PrintAligned(s, align); +} + +static void PrintDec(UInt64 v, size_t align) +{ + char s[32]; + ConvertUInt64ToString(v, s); + PrintAligned(s, align); +} + +static void PrintAddr(void *p) +{ + PrintHex((UInt64)(size_t)(ptrdiff_t)p, 12); +} + + +#define PRINT_ALLOC(name, cnt, size, ptr) \ + Print(name " "); \ + PrintDec(cnt++, 10); \ + PrintHex(size, 10); \ + PrintAddr(ptr); \ + PrintLn(); + +#define PRINT_FREE(name, cnt, ptr) if (ptr) { \ + Print(name " "); \ + PrintDec(--cnt, 10); \ + PrintAddr(ptr); \ + PrintLn(); } + +#else + +#define PRINT_ALLOC(name, cnt, size, ptr) +#define PRINT_FREE(name, cnt, ptr) +#define Print(s) +#define PrintLn() +#define PrintHex(v, align) +#define PrintAddr(p) + +#endif + + + +void *MyAlloc(size_t size) +{ + if (size == 0) + return NULL; + PRINT_ALLOC("Alloc ", g_allocCount, size, NULL); + #ifdef _SZ_ALLOC_DEBUG + { + void *p = malloc(size); + // PRINT_ALLOC("Alloc ", g_allocCount, size, p); + return p; + } + #else + return malloc(size); + #endif +} + +void MyFree(void *address) +{ + PRINT_FREE("Free ", g_allocCount, address); + + free(address); +} + +#ifdef _WIN32 + +void *MidAlloc(size_t size) +{ + if (size == 0) + return NULL; + + PRINT_ALLOC("Alloc-Mid", g_allocCountMid, size, NULL); + + return VirtualAlloc(NULL, size, MEM_COMMIT, PAGE_READWRITE); +} + +void MidFree(void *address) +{ + PRINT_FREE("Free-Mid", g_allocCountMid, address); + + if (!address) + return; + VirtualFree(address, 0, MEM_RELEASE); +} + +#ifdef _7ZIP_LARGE_PAGES + +#ifdef MEM_LARGE_PAGES + #define MY__MEM_LARGE_PAGES MEM_LARGE_PAGES +#else + #define MY__MEM_LARGE_PAGES 0x20000000 +#endif + +extern +SIZE_T g_LargePageSize; +SIZE_T g_LargePageSize = 0; +typedef SIZE_T (WINAPI *GetLargePageMinimumP)(VOID); + +#endif // _7ZIP_LARGE_PAGES + +void SetLargePageSize() +{ + #ifdef _7ZIP_LARGE_PAGES + SIZE_T size; + GetLargePageMinimumP largePageMinimum = (GetLargePageMinimumP) + GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "GetLargePageMinimum"); + if (!largePageMinimum) + return; + size = largePageMinimum(); + if (size == 0 || (size & (size - 1)) != 0) + return; + g_LargePageSize = size; + #endif +} + + +void *BigAlloc(size_t size) +{ + if (size == 0) + return NULL; + + PRINT_ALLOC("Alloc-Big", g_allocCountBig, size, NULL); + + #ifdef _7ZIP_LARGE_PAGES + { + SIZE_T ps = g_LargePageSize; + if (ps != 0 && ps <= (1 << 30) && size > (ps / 2)) + { + size_t size2; + ps--; + size2 = (size + ps) & ~ps; + if (size2 >= size) + { + void *res = VirtualAlloc(NULL, size2, MEM_COMMIT | MY__MEM_LARGE_PAGES, PAGE_READWRITE); + if (res) + return res; + } + } + } + #endif + + return VirtualAlloc(NULL, size, MEM_COMMIT, PAGE_READWRITE); +} + +void BigFree(void *address) +{ + PRINT_FREE("Free-Big", g_allocCountBig, address); + + if (!address) + return; + VirtualFree(address, 0, MEM_RELEASE); +} + +#endif + + +static void *SzAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return MyAlloc(size); } +static void SzFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); MyFree(address); } +const ISzAlloc g_Alloc = { SzAlloc, SzFree }; + +#ifdef _WIN32 +static void *SzMidAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return MidAlloc(size); } +static void SzMidFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); MidFree(address); } +static void *SzBigAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return BigAlloc(size); } +static void SzBigFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); BigFree(address); } +const ISzAlloc g_MidAlloc = { SzMidAlloc, SzMidFree }; +const ISzAlloc g_BigAlloc = { SzBigAlloc, SzBigFree }; +#endif + +/* + uintptr_t : C99 (optional) + : unsupported in VS6 +*/ + +#ifdef _WIN32 + typedef UINT_PTR UIntPtr; +#else + /* + typedef uintptr_t UIntPtr; + */ + typedef ptrdiff_t UIntPtr; +#endif + + +#define ADJUST_ALLOC_SIZE 0 +/* +#define ADJUST_ALLOC_SIZE (sizeof(void *) - 1) +*/ +/* + Use (ADJUST_ALLOC_SIZE = (sizeof(void *) - 1)), if + MyAlloc() can return address that is NOT multiple of sizeof(void *). +*/ + + +/* +#define MY_ALIGN_PTR_DOWN(p, align) ((void *)((char *)(p) - ((size_t)(UIntPtr)(p) & ((align) - 1)))) +*/ +#define MY_ALIGN_PTR_DOWN(p, align) ((void *)((((UIntPtr)(p)) & ~((UIntPtr)(align) - 1)))) + + +#if !defined(_WIN32) && defined(_POSIX_C_SOURCE) && (_POSIX_C_SOURCE >= 200112L) + #define USE_posix_memalign +#endif + +#ifndef USE_posix_memalign +#define MY_ALIGN_PTR_UP_PLUS(p, align) MY_ALIGN_PTR_DOWN(((char *)(p) + (align) + ADJUST_ALLOC_SIZE), align) +#endif + +/* + This posix_memalign() is for test purposes only. + We also need special Free() function instead of free(), + if this posix_memalign() is used. +*/ + +/* +static int posix_memalign(void **ptr, size_t align, size_t size) +{ + size_t newSize = size + align; + void *p; + void *pAligned; + *ptr = NULL; + if (newSize < size) + return 12; // ENOMEM + p = MyAlloc(newSize); + if (!p) + return 12; // ENOMEM + pAligned = MY_ALIGN_PTR_UP_PLUS(p, align); + ((void **)pAligned)[-1] = p; + *ptr = pAligned; + return 0; +} +*/ + +/* + ALLOC_ALIGN_SIZE >= sizeof(void *) + ALLOC_ALIGN_SIZE >= cache_line_size +*/ + +#define ALLOC_ALIGN_SIZE ((size_t)1 << 7) + +static void *SzAlignedAlloc(ISzAllocPtr pp, size_t size) +{ + #ifndef USE_posix_memalign + + void *p; + void *pAligned; + size_t newSize; + UNUSED_VAR(pp); + + /* also we can allocate additional dummy ALLOC_ALIGN_SIZE bytes after aligned + block to prevent cache line sharing with another allocated blocks */ + + newSize = size + ALLOC_ALIGN_SIZE * 1 + ADJUST_ALLOC_SIZE; + if (newSize < size) + return NULL; + + p = MyAlloc(newSize); + + if (!p) + return NULL; + pAligned = MY_ALIGN_PTR_UP_PLUS(p, ALLOC_ALIGN_SIZE); + + Print(" size="); PrintHex(size, 8); + Print(" a_size="); PrintHex(newSize, 8); + Print(" ptr="); PrintAddr(p); + Print(" a_ptr="); PrintAddr(pAligned); + PrintLn(); + + ((void **)pAligned)[-1] = p; + + return pAligned; + + #else + + void *p; + UNUSED_VAR(pp); + if (posix_memalign(&p, ALLOC_ALIGN_SIZE, size)) + return NULL; + + Print(" posix_memalign="); PrintAddr(p); + PrintLn(); + + return p; + + #endif +} + + +static void SzAlignedFree(ISzAllocPtr pp, void *address) +{ + UNUSED_VAR(pp); + #ifndef USE_posix_memalign + if (address) + MyFree(((void **)address)[-1]); + #else + free(address); + #endif +} + + +const ISzAlloc g_AlignedAlloc = { SzAlignedAlloc, SzAlignedFree }; + + + +#define MY_ALIGN_PTR_DOWN_1(p) MY_ALIGN_PTR_DOWN(p, sizeof(void *)) + +/* we align ptr to support cases where CAlignOffsetAlloc::offset is not multiply of sizeof(void *) */ +#define REAL_BLOCK_PTR_VAR(p) ((void **)MY_ALIGN_PTR_DOWN_1(p))[-1] +/* +#define REAL_BLOCK_PTR_VAR(p) ((void **)(p))[-1] +*/ + +static void *AlignOffsetAlloc_Alloc(ISzAllocPtr pp, size_t size) +{ + CAlignOffsetAlloc *p = CONTAINER_FROM_VTBL(pp, CAlignOffsetAlloc, vt); + void *adr; + void *pAligned; + size_t newSize; + size_t extra; + size_t alignSize = (size_t)1 << p->numAlignBits; + + if (alignSize < sizeof(void *)) + alignSize = sizeof(void *); + + if (p->offset >= alignSize) + return NULL; + + /* also we can allocate additional dummy ALLOC_ALIGN_SIZE bytes after aligned + block to prevent cache line sharing with another allocated blocks */ + extra = p->offset & (sizeof(void *) - 1); + newSize = size + alignSize + extra + ADJUST_ALLOC_SIZE; + if (newSize < size) + return NULL; + + adr = ISzAlloc_Alloc(p->baseAlloc, newSize); + + if (!adr) + return NULL; + + pAligned = (char *)MY_ALIGN_PTR_DOWN((char *)adr + + alignSize - p->offset + extra + ADJUST_ALLOC_SIZE, alignSize) + p->offset; + + PrintLn(); + Print("- Aligned: "); + Print(" size="); PrintHex(size, 8); + Print(" a_size="); PrintHex(newSize, 8); + Print(" ptr="); PrintAddr(adr); + Print(" a_ptr="); PrintAddr(pAligned); + PrintLn(); + + REAL_BLOCK_PTR_VAR(pAligned) = adr; + + return pAligned; +} + + +static void AlignOffsetAlloc_Free(ISzAllocPtr pp, void *address) +{ + if (address) + { + CAlignOffsetAlloc *p = CONTAINER_FROM_VTBL(pp, CAlignOffsetAlloc, vt); + PrintLn(); + Print("- Aligned Free: "); + PrintLn(); + ISzAlloc_Free(p->baseAlloc, REAL_BLOCK_PTR_VAR(address)); + } +} + + +void AlignOffsetAlloc_CreateVTable(CAlignOffsetAlloc *p) +{ + p->vt.Alloc = AlignOffsetAlloc_Alloc; + p->vt.Free = AlignOffsetAlloc_Free; +} diff --git a/deps/LZMA-SDK/C/Alloc.h b/deps/LZMA-SDK/C/Alloc.h index 59de10760..3be2041eb 100644 --- a/deps/LZMA-SDK/C/Alloc.h +++ b/deps/LZMA-SDK/C/Alloc.h @@ -1,58 +1,58 @@ -/* Alloc.h -- Memory allocation functions -2021-07-13 : Igor Pavlov : Public domain */ - -#ifndef __COMMON_ALLOC_H -#define __COMMON_ALLOC_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -void *MyAlloc(size_t size); -void MyFree(void *address); - -#ifdef _WIN32 - -void SetLargePageSize(void); - -void *MidAlloc(size_t size); -void MidFree(void *address); -void *BigAlloc(size_t size); -void BigFree(void *address); - -#else - -#define MidAlloc(size) MyAlloc(size) -#define MidFree(address) MyFree(address) -#define BigAlloc(size) MyAlloc(size) -#define BigFree(address) MyFree(address) - -#endif - -extern const ISzAlloc g_Alloc; - -#ifdef _WIN32 -extern const ISzAlloc g_BigAlloc; -extern const ISzAlloc g_MidAlloc; -#else -#define g_BigAlloc g_AlignedAlloc -#define g_MidAlloc g_AlignedAlloc -#endif - -extern const ISzAlloc g_AlignedAlloc; - - -typedef struct -{ - ISzAlloc vt; - ISzAllocPtr baseAlloc; - unsigned numAlignBits; /* ((1 << numAlignBits) >= sizeof(void *)) */ - size_t offset; /* (offset == (k * sizeof(void *)) && offset < (1 << numAlignBits) */ -} CAlignOffsetAlloc; - -void AlignOffsetAlloc_CreateVTable(CAlignOffsetAlloc *p); - - -EXTERN_C_END - -#endif +/* Alloc.h -- Memory allocation functions +2021-07-13 : Igor Pavlov : Public domain */ + +#ifndef __COMMON_ALLOC_H +#define __COMMON_ALLOC_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +void *MyAlloc(size_t size); +void MyFree(void *address); + +#ifdef _WIN32 + +void SetLargePageSize(void); + +void *MidAlloc(size_t size); +void MidFree(void *address); +void *BigAlloc(size_t size); +void BigFree(void *address); + +#else + +#define MidAlloc(size) MyAlloc(size) +#define MidFree(address) MyFree(address) +#define BigAlloc(size) MyAlloc(size) +#define BigFree(address) MyFree(address) + +#endif + +extern const ISzAlloc g_Alloc; + +#ifdef _WIN32 +extern const ISzAlloc g_BigAlloc; +extern const ISzAlloc g_MidAlloc; +#else +#define g_BigAlloc g_AlignedAlloc +#define g_MidAlloc g_AlignedAlloc +#endif + +extern const ISzAlloc g_AlignedAlloc; + + +typedef struct +{ + ISzAlloc vt; + ISzAllocPtr baseAlloc; + unsigned numAlignBits; /* ((1 << numAlignBits) >= sizeof(void *)) */ + size_t offset; /* (offset == (k * sizeof(void *)) && offset < (1 << numAlignBits) */ +} CAlignOffsetAlloc; + +void AlignOffsetAlloc_CreateVTable(CAlignOffsetAlloc *p); + + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/Bcj2.c b/deps/LZMA-SDK/C/Bcj2.c index c1772f234..c7b956708 100644 --- a/deps/LZMA-SDK/C/Bcj2.c +++ b/deps/LZMA-SDK/C/Bcj2.c @@ -1,257 +1,257 @@ -/* Bcj2.c -- BCJ2 Decoder (Converter for x86 code) -2021-02-09 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "Bcj2.h" -#include "CpuArch.h" - -#define CProb UInt16 - -#define kTopValue ((UInt32)1 << 24) -#define kNumModelBits 11 -#define kBitModelTotal (1 << kNumModelBits) -#define kNumMoveBits 5 - -#define _IF_BIT_0 ttt = *prob; bound = (p->range >> kNumModelBits) * ttt; if (p->code < bound) -#define _UPDATE_0 p->range = bound; *prob = (CProb)(ttt + ((kBitModelTotal - ttt) >> kNumMoveBits)); -#define _UPDATE_1 p->range -= bound; p->code -= bound; *prob = (CProb)(ttt - (ttt >> kNumMoveBits)); - -void Bcj2Dec_Init(CBcj2Dec *p) -{ - unsigned i; - - p->state = BCJ2_DEC_STATE_OK; - p->ip = 0; - p->temp[3] = 0; - p->range = 0; - p->code = 0; - for (i = 0; i < sizeof(p->probs) / sizeof(p->probs[0]); i++) - p->probs[i] = kBitModelTotal >> 1; -} - -SRes Bcj2Dec_Decode(CBcj2Dec *p) -{ - if (p->range <= 5) - { - p->state = BCJ2_DEC_STATE_OK; - for (; p->range != 5; p->range++) - { - if (p->range == 1 && p->code != 0) - return SZ_ERROR_DATA; - - if (p->bufs[BCJ2_STREAM_RC] == p->lims[BCJ2_STREAM_RC]) - { - p->state = BCJ2_STREAM_RC; - return SZ_OK; - } - - p->code = (p->code << 8) | *(p->bufs[BCJ2_STREAM_RC])++; - } - - if (p->code == 0xFFFFFFFF) - return SZ_ERROR_DATA; - - p->range = 0xFFFFFFFF; - } - else if (p->state >= BCJ2_DEC_STATE_ORIG_0) - { - while (p->state <= BCJ2_DEC_STATE_ORIG_3) - { - Byte *dest = p->dest; - if (dest == p->destLim) - return SZ_OK; - *dest = p->temp[(size_t)p->state - BCJ2_DEC_STATE_ORIG_0]; - p->state++; - p->dest = dest + 1; - } - } - - /* - if (BCJ2_IS_32BIT_STREAM(p->state)) - { - const Byte *cur = p->bufs[p->state]; - if (cur == p->lims[p->state]) - return SZ_OK; - p->bufs[p->state] = cur + 4; - - { - UInt32 val; - Byte *dest; - SizeT rem; - - p->ip += 4; - val = GetBe32(cur) - p->ip; - dest = p->dest; - rem = p->destLim - dest; - if (rem < 4) - { - SizeT i; - SetUi32(p->temp, val); - for (i = 0; i < rem; i++) - dest[i] = p->temp[i]; - p->dest = dest + rem; - p->state = BCJ2_DEC_STATE_ORIG_0 + (unsigned)rem; - return SZ_OK; - } - SetUi32(dest, val); - p->temp[3] = (Byte)(val >> 24); - p->dest = dest + 4; - p->state = BCJ2_DEC_STATE_OK; - } - } - */ - - for (;;) - { - if (BCJ2_IS_32BIT_STREAM(p->state)) - p->state = BCJ2_DEC_STATE_OK; - else - { - if (p->range < kTopValue) - { - if (p->bufs[BCJ2_STREAM_RC] == p->lims[BCJ2_STREAM_RC]) - { - p->state = BCJ2_STREAM_RC; - return SZ_OK; - } - p->range <<= 8; - p->code = (p->code << 8) | *(p->bufs[BCJ2_STREAM_RC])++; - } - - { - const Byte *src = p->bufs[BCJ2_STREAM_MAIN]; - const Byte *srcLim; - Byte *dest; - SizeT num = (SizeT)(p->lims[BCJ2_STREAM_MAIN] - src); - - if (num == 0) - { - p->state = BCJ2_STREAM_MAIN; - return SZ_OK; - } - - dest = p->dest; - if (num > (SizeT)(p->destLim - dest)) - { - num = (SizeT)(p->destLim - dest); - if (num == 0) - { - p->state = BCJ2_DEC_STATE_ORIG; - return SZ_OK; - } - } - - srcLim = src + num; - - if (p->temp[3] == 0x0F && (src[0] & 0xF0) == 0x80) - *dest = src[0]; - else for (;;) - { - Byte b = *src; - *dest = b; - if (b != 0x0F) - { - if ((b & 0xFE) == 0xE8) - break; - dest++; - if (++src != srcLim) - continue; - break; - } - dest++; - if (++src == srcLim) - break; - if ((*src & 0xF0) != 0x80) - continue; - *dest = *src; - break; - } - - num = (SizeT)(src - p->bufs[BCJ2_STREAM_MAIN]); - - if (src == srcLim) - { - p->temp[3] = src[-1]; - p->bufs[BCJ2_STREAM_MAIN] = src; - p->ip += (UInt32)num; - p->dest += num; - p->state = - p->bufs[BCJ2_STREAM_MAIN] == - p->lims[BCJ2_STREAM_MAIN] ? - (unsigned)BCJ2_STREAM_MAIN : - (unsigned)BCJ2_DEC_STATE_ORIG; - return SZ_OK; - } - - { - UInt32 bound, ttt; - CProb *prob; - Byte b = src[0]; - Byte prev = (Byte)(num == 0 ? p->temp[3] : src[-1]); - - p->temp[3] = b; - p->bufs[BCJ2_STREAM_MAIN] = src + 1; - num++; - p->ip += (UInt32)num; - p->dest += num; - - prob = p->probs + (unsigned)(b == 0xE8 ? 2 + (unsigned)prev : (b == 0xE9 ? 1 : 0)); - - _IF_BIT_0 - { - _UPDATE_0 - continue; - } - _UPDATE_1 - - } - } - } - - { - UInt32 val; - unsigned cj = (p->temp[3] == 0xE8) ? BCJ2_STREAM_CALL : BCJ2_STREAM_JUMP; - const Byte *cur = p->bufs[cj]; - Byte *dest; - SizeT rem; - - if (cur == p->lims[cj]) - { - p->state = cj; - break; - } - - val = GetBe32(cur); - p->bufs[cj] = cur + 4; - - p->ip += 4; - val -= p->ip; - dest = p->dest; - rem = (SizeT)(p->destLim - dest); - - if (rem < 4) - { - p->temp[0] = (Byte)val; if (rem > 0) dest[0] = (Byte)val; val >>= 8; - p->temp[1] = (Byte)val; if (rem > 1) dest[1] = (Byte)val; val >>= 8; - p->temp[2] = (Byte)val; if (rem > 2) dest[2] = (Byte)val; val >>= 8; - p->temp[3] = (Byte)val; - p->dest = dest + rem; - p->state = BCJ2_DEC_STATE_ORIG_0 + (unsigned)rem; - break; - } - - SetUi32(dest, val); - p->temp[3] = (Byte)(val >> 24); - p->dest = dest + 4; - } - } - - if (p->range < kTopValue && p->bufs[BCJ2_STREAM_RC] != p->lims[BCJ2_STREAM_RC]) - { - p->range <<= 8; - p->code = (p->code << 8) | *(p->bufs[BCJ2_STREAM_RC])++; - } - - return SZ_OK; -} +/* Bcj2.c -- BCJ2 Decoder (Converter for x86 code) +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "Bcj2.h" +#include "CpuArch.h" + +#define CProb UInt16 + +#define kTopValue ((UInt32)1 << 24) +#define kNumModelBits 11 +#define kBitModelTotal (1 << kNumModelBits) +#define kNumMoveBits 5 + +#define _IF_BIT_0 ttt = *prob; bound = (p->range >> kNumModelBits) * ttt; if (p->code < bound) +#define _UPDATE_0 p->range = bound; *prob = (CProb)(ttt + ((kBitModelTotal - ttt) >> kNumMoveBits)); +#define _UPDATE_1 p->range -= bound; p->code -= bound; *prob = (CProb)(ttt - (ttt >> kNumMoveBits)); + +void Bcj2Dec_Init(CBcj2Dec *p) +{ + unsigned i; + + p->state = BCJ2_DEC_STATE_OK; + p->ip = 0; + p->temp[3] = 0; + p->range = 0; + p->code = 0; + for (i = 0; i < sizeof(p->probs) / sizeof(p->probs[0]); i++) + p->probs[i] = kBitModelTotal >> 1; +} + +SRes Bcj2Dec_Decode(CBcj2Dec *p) +{ + if (p->range <= 5) + { + p->state = BCJ2_DEC_STATE_OK; + for (; p->range != 5; p->range++) + { + if (p->range == 1 && p->code != 0) + return SZ_ERROR_DATA; + + if (p->bufs[BCJ2_STREAM_RC] == p->lims[BCJ2_STREAM_RC]) + { + p->state = BCJ2_STREAM_RC; + return SZ_OK; + } + + p->code = (p->code << 8) | *(p->bufs[BCJ2_STREAM_RC])++; + } + + if (p->code == 0xFFFFFFFF) + return SZ_ERROR_DATA; + + p->range = 0xFFFFFFFF; + } + else if (p->state >= BCJ2_DEC_STATE_ORIG_0) + { + while (p->state <= BCJ2_DEC_STATE_ORIG_3) + { + Byte *dest = p->dest; + if (dest == p->destLim) + return SZ_OK; + *dest = p->temp[(size_t)p->state - BCJ2_DEC_STATE_ORIG_0]; + p->state++; + p->dest = dest + 1; + } + } + + /* + if (BCJ2_IS_32BIT_STREAM(p->state)) + { + const Byte *cur = p->bufs[p->state]; + if (cur == p->lims[p->state]) + return SZ_OK; + p->bufs[p->state] = cur + 4; + + { + UInt32 val; + Byte *dest; + SizeT rem; + + p->ip += 4; + val = GetBe32(cur) - p->ip; + dest = p->dest; + rem = p->destLim - dest; + if (rem < 4) + { + SizeT i; + SetUi32(p->temp, val); + for (i = 0; i < rem; i++) + dest[i] = p->temp[i]; + p->dest = dest + rem; + p->state = BCJ2_DEC_STATE_ORIG_0 + (unsigned)rem; + return SZ_OK; + } + SetUi32(dest, val); + p->temp[3] = (Byte)(val >> 24); + p->dest = dest + 4; + p->state = BCJ2_DEC_STATE_OK; + } + } + */ + + for (;;) + { + if (BCJ2_IS_32BIT_STREAM(p->state)) + p->state = BCJ2_DEC_STATE_OK; + else + { + if (p->range < kTopValue) + { + if (p->bufs[BCJ2_STREAM_RC] == p->lims[BCJ2_STREAM_RC]) + { + p->state = BCJ2_STREAM_RC; + return SZ_OK; + } + p->range <<= 8; + p->code = (p->code << 8) | *(p->bufs[BCJ2_STREAM_RC])++; + } + + { + const Byte *src = p->bufs[BCJ2_STREAM_MAIN]; + const Byte *srcLim; + Byte *dest; + SizeT num = (SizeT)(p->lims[BCJ2_STREAM_MAIN] - src); + + if (num == 0) + { + p->state = BCJ2_STREAM_MAIN; + return SZ_OK; + } + + dest = p->dest; + if (num > (SizeT)(p->destLim - dest)) + { + num = (SizeT)(p->destLim - dest); + if (num == 0) + { + p->state = BCJ2_DEC_STATE_ORIG; + return SZ_OK; + } + } + + srcLim = src + num; + + if (p->temp[3] == 0x0F && (src[0] & 0xF0) == 0x80) + *dest = src[0]; + else for (;;) + { + Byte b = *src; + *dest = b; + if (b != 0x0F) + { + if ((b & 0xFE) == 0xE8) + break; + dest++; + if (++src != srcLim) + continue; + break; + } + dest++; + if (++src == srcLim) + break; + if ((*src & 0xF0) != 0x80) + continue; + *dest = *src; + break; + } + + num = (SizeT)(src - p->bufs[BCJ2_STREAM_MAIN]); + + if (src == srcLim) + { + p->temp[3] = src[-1]; + p->bufs[BCJ2_STREAM_MAIN] = src; + p->ip += (UInt32)num; + p->dest += num; + p->state = + p->bufs[BCJ2_STREAM_MAIN] == + p->lims[BCJ2_STREAM_MAIN] ? + (unsigned)BCJ2_STREAM_MAIN : + (unsigned)BCJ2_DEC_STATE_ORIG; + return SZ_OK; + } + + { + UInt32 bound, ttt; + CProb *prob; + Byte b = src[0]; + Byte prev = (Byte)(num == 0 ? p->temp[3] : src[-1]); + + p->temp[3] = b; + p->bufs[BCJ2_STREAM_MAIN] = src + 1; + num++; + p->ip += (UInt32)num; + p->dest += num; + + prob = p->probs + (unsigned)(b == 0xE8 ? 2 + (unsigned)prev : (b == 0xE9 ? 1 : 0)); + + _IF_BIT_0 + { + _UPDATE_0 + continue; + } + _UPDATE_1 + + } + } + } + + { + UInt32 val; + unsigned cj = (p->temp[3] == 0xE8) ? BCJ2_STREAM_CALL : BCJ2_STREAM_JUMP; + const Byte *cur = p->bufs[cj]; + Byte *dest; + SizeT rem; + + if (cur == p->lims[cj]) + { + p->state = cj; + break; + } + + val = GetBe32(cur); + p->bufs[cj] = cur + 4; + + p->ip += 4; + val -= p->ip; + dest = p->dest; + rem = (SizeT)(p->destLim - dest); + + if (rem < 4) + { + p->temp[0] = (Byte)val; if (rem > 0) dest[0] = (Byte)val; val >>= 8; + p->temp[1] = (Byte)val; if (rem > 1) dest[1] = (Byte)val; val >>= 8; + p->temp[2] = (Byte)val; if (rem > 2) dest[2] = (Byte)val; val >>= 8; + p->temp[3] = (Byte)val; + p->dest = dest + rem; + p->state = BCJ2_DEC_STATE_ORIG_0 + (unsigned)rem; + break; + } + + SetUi32(dest, val); + p->temp[3] = (Byte)(val >> 24); + p->dest = dest + 4; + } + } + + if (p->range < kTopValue && p->bufs[BCJ2_STREAM_RC] != p->lims[BCJ2_STREAM_RC]) + { + p->range <<= 8; + p->code = (p->code << 8) | *(p->bufs[BCJ2_STREAM_RC])++; + } + + return SZ_OK; +} diff --git a/deps/LZMA-SDK/C/Bcj2.h b/deps/LZMA-SDK/C/Bcj2.h index 68893d2d1..8824080ac 100644 --- a/deps/LZMA-SDK/C/Bcj2.h +++ b/deps/LZMA-SDK/C/Bcj2.h @@ -1,146 +1,146 @@ -/* Bcj2.h -- BCJ2 Converter for x86 code -2014-11-10 : Igor Pavlov : Public domain */ - -#ifndef __BCJ2_H -#define __BCJ2_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -#define BCJ2_NUM_STREAMS 4 - -enum -{ - BCJ2_STREAM_MAIN, - BCJ2_STREAM_CALL, - BCJ2_STREAM_JUMP, - BCJ2_STREAM_RC -}; - -enum -{ - BCJ2_DEC_STATE_ORIG_0 = BCJ2_NUM_STREAMS, - BCJ2_DEC_STATE_ORIG_1, - BCJ2_DEC_STATE_ORIG_2, - BCJ2_DEC_STATE_ORIG_3, - - BCJ2_DEC_STATE_ORIG, - BCJ2_DEC_STATE_OK -}; - -enum -{ - BCJ2_ENC_STATE_ORIG = BCJ2_NUM_STREAMS, - BCJ2_ENC_STATE_OK -}; - - -#define BCJ2_IS_32BIT_STREAM(s) ((s) == BCJ2_STREAM_CALL || (s) == BCJ2_STREAM_JUMP) - -/* -CBcj2Dec / CBcj2Enc -bufs sizes: - BUF_SIZE(n) = lims[n] - bufs[n] -bufs sizes for BCJ2_STREAM_CALL and BCJ2_STREAM_JUMP must be mutliply of 4: - (BUF_SIZE(BCJ2_STREAM_CALL) & 3) == 0 - (BUF_SIZE(BCJ2_STREAM_JUMP) & 3) == 0 -*/ - -/* -CBcj2Dec: -dest is allowed to overlap with bufs[BCJ2_STREAM_MAIN], with the following conditions: - bufs[BCJ2_STREAM_MAIN] >= dest && - bufs[BCJ2_STREAM_MAIN] - dest >= tempReserv + - BUF_SIZE(BCJ2_STREAM_CALL) + - BUF_SIZE(BCJ2_STREAM_JUMP) - tempReserv = 0 : for first call of Bcj2Dec_Decode - tempReserv = 4 : for any other calls of Bcj2Dec_Decode - overlap with offset = 1 is not allowed -*/ - -typedef struct -{ - const Byte *bufs[BCJ2_NUM_STREAMS]; - const Byte *lims[BCJ2_NUM_STREAMS]; - Byte *dest; - const Byte *destLim; - - unsigned state; /* BCJ2_STREAM_MAIN has more priority than BCJ2_STATE_ORIG */ - - UInt32 ip; - Byte temp[4]; - UInt32 range; - UInt32 code; - UInt16 probs[2 + 256]; -} CBcj2Dec; - -void Bcj2Dec_Init(CBcj2Dec *p); - -/* Returns: SZ_OK or SZ_ERROR_DATA */ -SRes Bcj2Dec_Decode(CBcj2Dec *p); - -#define Bcj2Dec_IsFinished(_p_) ((_p_)->code == 0) - - - -typedef enum -{ - BCJ2_ENC_FINISH_MODE_CONTINUE, - BCJ2_ENC_FINISH_MODE_END_BLOCK, - BCJ2_ENC_FINISH_MODE_END_STREAM -} EBcj2Enc_FinishMode; - -typedef struct -{ - Byte *bufs[BCJ2_NUM_STREAMS]; - const Byte *lims[BCJ2_NUM_STREAMS]; - const Byte *src; - const Byte *srcLim; - - unsigned state; - EBcj2Enc_FinishMode finishMode; - - Byte prevByte; - - Byte cache; - UInt32 range; - UInt64 low; - UInt64 cacheSize; - - UInt32 ip; - - /* 32-bit ralative offset in JUMP/CALL commands is - - (mod 4 GB) in 32-bit mode - - signed Int32 in 64-bit mode - We use (mod 4 GB) check for fileSize. - Use fileSize up to 2 GB, if you want to support 32-bit and 64-bit code conversion. */ - UInt32 fileIp; - UInt32 fileSize; /* (fileSize <= ((UInt32)1 << 31)), 0 means no_limit */ - UInt32 relatLimit; /* (relatLimit <= ((UInt32)1 << 31)), 0 means desable_conversion */ - - UInt32 tempTarget; - unsigned tempPos; - Byte temp[4 * 2]; - - unsigned flushPos; - - UInt16 probs[2 + 256]; -} CBcj2Enc; - -void Bcj2Enc_Init(CBcj2Enc *p); -void Bcj2Enc_Encode(CBcj2Enc *p); - -#define Bcj2Enc_Get_InputData_Size(p) ((SizeT)((p)->srcLim - (p)->src) + (p)->tempPos) -#define Bcj2Enc_IsFinished(p) ((p)->flushPos == 5) - - -#define BCJ2_RELAT_LIMIT_NUM_BITS 26 -#define BCJ2_RELAT_LIMIT ((UInt32)1 << BCJ2_RELAT_LIMIT_NUM_BITS) - -/* limit for CBcj2Enc::fileSize variable */ -#define BCJ2_FileSize_MAX ((UInt32)1 << 31) - -EXTERN_C_END - -#endif +/* Bcj2.h -- BCJ2 Converter for x86 code +2014-11-10 : Igor Pavlov : Public domain */ + +#ifndef __BCJ2_H +#define __BCJ2_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define BCJ2_NUM_STREAMS 4 + +enum +{ + BCJ2_STREAM_MAIN, + BCJ2_STREAM_CALL, + BCJ2_STREAM_JUMP, + BCJ2_STREAM_RC +}; + +enum +{ + BCJ2_DEC_STATE_ORIG_0 = BCJ2_NUM_STREAMS, + BCJ2_DEC_STATE_ORIG_1, + BCJ2_DEC_STATE_ORIG_2, + BCJ2_DEC_STATE_ORIG_3, + + BCJ2_DEC_STATE_ORIG, + BCJ2_DEC_STATE_OK +}; + +enum +{ + BCJ2_ENC_STATE_ORIG = BCJ2_NUM_STREAMS, + BCJ2_ENC_STATE_OK +}; + + +#define BCJ2_IS_32BIT_STREAM(s) ((s) == BCJ2_STREAM_CALL || (s) == BCJ2_STREAM_JUMP) + +/* +CBcj2Dec / CBcj2Enc +bufs sizes: + BUF_SIZE(n) = lims[n] - bufs[n] +bufs sizes for BCJ2_STREAM_CALL and BCJ2_STREAM_JUMP must be mutliply of 4: + (BUF_SIZE(BCJ2_STREAM_CALL) & 3) == 0 + (BUF_SIZE(BCJ2_STREAM_JUMP) & 3) == 0 +*/ + +/* +CBcj2Dec: +dest is allowed to overlap with bufs[BCJ2_STREAM_MAIN], with the following conditions: + bufs[BCJ2_STREAM_MAIN] >= dest && + bufs[BCJ2_STREAM_MAIN] - dest >= tempReserv + + BUF_SIZE(BCJ2_STREAM_CALL) + + BUF_SIZE(BCJ2_STREAM_JUMP) + tempReserv = 0 : for first call of Bcj2Dec_Decode + tempReserv = 4 : for any other calls of Bcj2Dec_Decode + overlap with offset = 1 is not allowed +*/ + +typedef struct +{ + const Byte *bufs[BCJ2_NUM_STREAMS]; + const Byte *lims[BCJ2_NUM_STREAMS]; + Byte *dest; + const Byte *destLim; + + unsigned state; /* BCJ2_STREAM_MAIN has more priority than BCJ2_STATE_ORIG */ + + UInt32 ip; + Byte temp[4]; + UInt32 range; + UInt32 code; + UInt16 probs[2 + 256]; +} CBcj2Dec; + +void Bcj2Dec_Init(CBcj2Dec *p); + +/* Returns: SZ_OK or SZ_ERROR_DATA */ +SRes Bcj2Dec_Decode(CBcj2Dec *p); + +#define Bcj2Dec_IsFinished(_p_) ((_p_)->code == 0) + + + +typedef enum +{ + BCJ2_ENC_FINISH_MODE_CONTINUE, + BCJ2_ENC_FINISH_MODE_END_BLOCK, + BCJ2_ENC_FINISH_MODE_END_STREAM +} EBcj2Enc_FinishMode; + +typedef struct +{ + Byte *bufs[BCJ2_NUM_STREAMS]; + const Byte *lims[BCJ2_NUM_STREAMS]; + const Byte *src; + const Byte *srcLim; + + unsigned state; + EBcj2Enc_FinishMode finishMode; + + Byte prevByte; + + Byte cache; + UInt32 range; + UInt64 low; + UInt64 cacheSize; + + UInt32 ip; + + /* 32-bit ralative offset in JUMP/CALL commands is + - (mod 4 GB) in 32-bit mode + - signed Int32 in 64-bit mode + We use (mod 4 GB) check for fileSize. + Use fileSize up to 2 GB, if you want to support 32-bit and 64-bit code conversion. */ + UInt32 fileIp; + UInt32 fileSize; /* (fileSize <= ((UInt32)1 << 31)), 0 means no_limit */ + UInt32 relatLimit; /* (relatLimit <= ((UInt32)1 << 31)), 0 means desable_conversion */ + + UInt32 tempTarget; + unsigned tempPos; + Byte temp[4 * 2]; + + unsigned flushPos; + + UInt16 probs[2 + 256]; +} CBcj2Enc; + +void Bcj2Enc_Init(CBcj2Enc *p); +void Bcj2Enc_Encode(CBcj2Enc *p); + +#define Bcj2Enc_Get_InputData_Size(p) ((SizeT)((p)->srcLim - (p)->src) + (p)->tempPos) +#define Bcj2Enc_IsFinished(p) ((p)->flushPos == 5) + + +#define BCJ2_RELAT_LIMIT_NUM_BITS 26 +#define BCJ2_RELAT_LIMIT ((UInt32)1 << BCJ2_RELAT_LIMIT_NUM_BITS) + +/* limit for CBcj2Enc::fileSize variable */ +#define BCJ2_FileSize_MAX ((UInt32)1 << 31) + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/Bcj2Enc.c b/deps/LZMA-SDK/C/Bcj2Enc.c index 71ac5091d..682362a15 100644 --- a/deps/LZMA-SDK/C/Bcj2Enc.c +++ b/deps/LZMA-SDK/C/Bcj2Enc.c @@ -1,311 +1,311 @@ -/* Bcj2Enc.c -- BCJ2 Encoder (Converter for x86 code) -2021-02-09 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -/* #define SHOW_STAT */ - -#ifdef SHOW_STAT -#include -#define PRF(x) x -#else -#define PRF(x) -#endif - -#include - -#include "Bcj2.h" -#include "CpuArch.h" - -#define CProb UInt16 - -#define kTopValue ((UInt32)1 << 24) -#define kNumModelBits 11 -#define kBitModelTotal (1 << kNumModelBits) -#define kNumMoveBits 5 - -void Bcj2Enc_Init(CBcj2Enc *p) -{ - unsigned i; - - p->state = BCJ2_ENC_STATE_OK; - p->finishMode = BCJ2_ENC_FINISH_MODE_CONTINUE; - - p->prevByte = 0; - - p->cache = 0; - p->range = 0xFFFFFFFF; - p->low = 0; - p->cacheSize = 1; - - p->ip = 0; - - p->fileIp = 0; - p->fileSize = 0; - p->relatLimit = BCJ2_RELAT_LIMIT; - - p->tempPos = 0; - - p->flushPos = 0; - - for (i = 0; i < sizeof(p->probs) / sizeof(p->probs[0]); i++) - p->probs[i] = kBitModelTotal >> 1; -} - -static BoolInt MY_FAST_CALL RangeEnc_ShiftLow(CBcj2Enc *p) -{ - if ((UInt32)p->low < (UInt32)0xFF000000 || (UInt32)(p->low >> 32) != 0) - { - Byte *buf = p->bufs[BCJ2_STREAM_RC]; - do - { - if (buf == p->lims[BCJ2_STREAM_RC]) - { - p->state = BCJ2_STREAM_RC; - p->bufs[BCJ2_STREAM_RC] = buf; - return True; - } - *buf++ = (Byte)(p->cache + (Byte)(p->low >> 32)); - p->cache = 0xFF; - } - while (--p->cacheSize); - p->bufs[BCJ2_STREAM_RC] = buf; - p->cache = (Byte)((UInt32)p->low >> 24); - } - p->cacheSize++; - p->low = (UInt32)p->low << 8; - return False; -} - -static void Bcj2Enc_Encode_2(CBcj2Enc *p) -{ - if (BCJ2_IS_32BIT_STREAM(p->state)) - { - Byte *cur = p->bufs[p->state]; - if (cur == p->lims[p->state]) - return; - SetBe32(cur, p->tempTarget); - p->bufs[p->state] = cur + 4; - } - - p->state = BCJ2_ENC_STATE_ORIG; - - for (;;) - { - if (p->range < kTopValue) - { - if (RangeEnc_ShiftLow(p)) - return; - p->range <<= 8; - } - - { - { - const Byte *src = p->src; - const Byte *srcLim; - Byte *dest; - SizeT num = (SizeT)(p->srcLim - src); - - if (p->finishMode == BCJ2_ENC_FINISH_MODE_CONTINUE) - { - if (num <= 4) - return; - num -= 4; - } - else if (num == 0) - break; - - dest = p->bufs[BCJ2_STREAM_MAIN]; - if (num > (SizeT)(p->lims[BCJ2_STREAM_MAIN] - dest)) - { - num = (SizeT)(p->lims[BCJ2_STREAM_MAIN] - dest); - if (num == 0) - { - p->state = BCJ2_STREAM_MAIN; - return; - } - } - - srcLim = src + num; - - if (p->prevByte == 0x0F && (src[0] & 0xF0) == 0x80) - *dest = src[0]; - else for (;;) - { - Byte b = *src; - *dest = b; - if (b != 0x0F) - { - if ((b & 0xFE) == 0xE8) - break; - dest++; - if (++src != srcLim) - continue; - break; - } - dest++; - if (++src == srcLim) - break; - if ((*src & 0xF0) != 0x80) - continue; - *dest = *src; - break; - } - - num = (SizeT)(src - p->src); - - if (src == srcLim) - { - p->prevByte = src[-1]; - p->bufs[BCJ2_STREAM_MAIN] = dest; - p->src = src; - p->ip += (UInt32)num; - continue; - } - - { - Byte context = (Byte)(num == 0 ? p->prevByte : src[-1]); - BoolInt needConvert; - - p->bufs[BCJ2_STREAM_MAIN] = dest + 1; - p->ip += (UInt32)num + 1; - src++; - - needConvert = False; - - if ((SizeT)(p->srcLim - src) >= 4) - { - UInt32 relatVal = GetUi32(src); - if ((p->fileSize == 0 || (UInt32)(p->ip + 4 + relatVal - p->fileIp) < p->fileSize) - && ((relatVal + p->relatLimit) >> 1) < p->relatLimit) - needConvert = True; - } - - { - UInt32 bound; - unsigned ttt; - Byte b = src[-1]; - CProb *prob = p->probs + (unsigned)(b == 0xE8 ? 2 + (unsigned)context : (b == 0xE9 ? 1 : 0)); - - ttt = *prob; - bound = (p->range >> kNumModelBits) * ttt; - - if (!needConvert) - { - p->range = bound; - *prob = (CProb)(ttt + ((kBitModelTotal - ttt) >> kNumMoveBits)); - p->src = src; - p->prevByte = b; - continue; - } - - p->low += bound; - p->range -= bound; - *prob = (CProb)(ttt - (ttt >> kNumMoveBits)); - - { - UInt32 relatVal = GetUi32(src); - UInt32 absVal; - p->ip += 4; - absVal = p->ip + relatVal; - p->prevByte = src[3]; - src += 4; - p->src = src; - { - unsigned cj = (b == 0xE8) ? BCJ2_STREAM_CALL : BCJ2_STREAM_JUMP; - Byte *cur = p->bufs[cj]; - if (cur == p->lims[cj]) - { - p->state = cj; - p->tempTarget = absVal; - return; - } - SetBe32(cur, absVal); - p->bufs[cj] = cur + 4; - } - } - } - } - } - } - } - - if (p->finishMode != BCJ2_ENC_FINISH_MODE_END_STREAM) - return; - - for (; p->flushPos < 5; p->flushPos++) - if (RangeEnc_ShiftLow(p)) - return; - p->state = BCJ2_ENC_STATE_OK; -} - - -void Bcj2Enc_Encode(CBcj2Enc *p) -{ - PRF(printf("\n")); - PRF(printf("---- ip = %8d tempPos = %8d src = %8d\n", p->ip, p->tempPos, p->srcLim - p->src)); - - if (p->tempPos != 0) - { - unsigned extra = 0; - - for (;;) - { - const Byte *src = p->src; - const Byte *srcLim = p->srcLim; - EBcj2Enc_FinishMode finishMode = p->finishMode; - - p->src = p->temp; - p->srcLim = p->temp + p->tempPos; - if (src != srcLim) - p->finishMode = BCJ2_ENC_FINISH_MODE_CONTINUE; - - PRF(printf(" ip = %8d tempPos = %8d src = %8d\n", p->ip, p->tempPos, p->srcLim - p->src)); - - Bcj2Enc_Encode_2(p); - - { - unsigned num = (unsigned)(p->src - p->temp); - unsigned tempPos = p->tempPos - num; - unsigned i; - p->tempPos = tempPos; - for (i = 0; i < tempPos; i++) - p->temp[i] = p->temp[(size_t)i + num]; - - p->src = src; - p->srcLim = srcLim; - p->finishMode = finishMode; - - if (p->state != BCJ2_ENC_STATE_ORIG || src == srcLim) - return; - - if (extra >= tempPos) - { - p->src = src - tempPos; - p->tempPos = 0; - break; - } - - p->temp[tempPos] = src[0]; - p->tempPos = tempPos + 1; - p->src = src + 1; - extra++; - } - } - } - - PRF(printf("++++ ip = %8d tempPos = %8d src = %8d\n", p->ip, p->tempPos, p->srcLim - p->src)); - - Bcj2Enc_Encode_2(p); - - if (p->state == BCJ2_ENC_STATE_ORIG) - { - const Byte *src = p->src; - unsigned rem = (unsigned)(p->srcLim - src); - unsigned i; - for (i = 0; i < rem; i++) - p->temp[i] = src[i]; - p->tempPos = rem; - p->src = src + rem; - } -} +/* Bcj2Enc.c -- BCJ2 Encoder (Converter for x86 code) +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +/* #define SHOW_STAT */ + +#ifdef SHOW_STAT +#include +#define PRF(x) x +#else +#define PRF(x) +#endif + +#include + +#include "Bcj2.h" +#include "CpuArch.h" + +#define CProb UInt16 + +#define kTopValue ((UInt32)1 << 24) +#define kNumModelBits 11 +#define kBitModelTotal (1 << kNumModelBits) +#define kNumMoveBits 5 + +void Bcj2Enc_Init(CBcj2Enc *p) +{ + unsigned i; + + p->state = BCJ2_ENC_STATE_OK; + p->finishMode = BCJ2_ENC_FINISH_MODE_CONTINUE; + + p->prevByte = 0; + + p->cache = 0; + p->range = 0xFFFFFFFF; + p->low = 0; + p->cacheSize = 1; + + p->ip = 0; + + p->fileIp = 0; + p->fileSize = 0; + p->relatLimit = BCJ2_RELAT_LIMIT; + + p->tempPos = 0; + + p->flushPos = 0; + + for (i = 0; i < sizeof(p->probs) / sizeof(p->probs[0]); i++) + p->probs[i] = kBitModelTotal >> 1; +} + +static BoolInt MY_FAST_CALL RangeEnc_ShiftLow(CBcj2Enc *p) +{ + if ((UInt32)p->low < (UInt32)0xFF000000 || (UInt32)(p->low >> 32) != 0) + { + Byte *buf = p->bufs[BCJ2_STREAM_RC]; + do + { + if (buf == p->lims[BCJ2_STREAM_RC]) + { + p->state = BCJ2_STREAM_RC; + p->bufs[BCJ2_STREAM_RC] = buf; + return True; + } + *buf++ = (Byte)(p->cache + (Byte)(p->low >> 32)); + p->cache = 0xFF; + } + while (--p->cacheSize); + p->bufs[BCJ2_STREAM_RC] = buf; + p->cache = (Byte)((UInt32)p->low >> 24); + } + p->cacheSize++; + p->low = (UInt32)p->low << 8; + return False; +} + +static void Bcj2Enc_Encode_2(CBcj2Enc *p) +{ + if (BCJ2_IS_32BIT_STREAM(p->state)) + { + Byte *cur = p->bufs[p->state]; + if (cur == p->lims[p->state]) + return; + SetBe32(cur, p->tempTarget); + p->bufs[p->state] = cur + 4; + } + + p->state = BCJ2_ENC_STATE_ORIG; + + for (;;) + { + if (p->range < kTopValue) + { + if (RangeEnc_ShiftLow(p)) + return; + p->range <<= 8; + } + + { + { + const Byte *src = p->src; + const Byte *srcLim; + Byte *dest; + SizeT num = (SizeT)(p->srcLim - src); + + if (p->finishMode == BCJ2_ENC_FINISH_MODE_CONTINUE) + { + if (num <= 4) + return; + num -= 4; + } + else if (num == 0) + break; + + dest = p->bufs[BCJ2_STREAM_MAIN]; + if (num > (SizeT)(p->lims[BCJ2_STREAM_MAIN] - dest)) + { + num = (SizeT)(p->lims[BCJ2_STREAM_MAIN] - dest); + if (num == 0) + { + p->state = BCJ2_STREAM_MAIN; + return; + } + } + + srcLim = src + num; + + if (p->prevByte == 0x0F && (src[0] & 0xF0) == 0x80) + *dest = src[0]; + else for (;;) + { + Byte b = *src; + *dest = b; + if (b != 0x0F) + { + if ((b & 0xFE) == 0xE8) + break; + dest++; + if (++src != srcLim) + continue; + break; + } + dest++; + if (++src == srcLim) + break; + if ((*src & 0xF0) != 0x80) + continue; + *dest = *src; + break; + } + + num = (SizeT)(src - p->src); + + if (src == srcLim) + { + p->prevByte = src[-1]; + p->bufs[BCJ2_STREAM_MAIN] = dest; + p->src = src; + p->ip += (UInt32)num; + continue; + } + + { + Byte context = (Byte)(num == 0 ? p->prevByte : src[-1]); + BoolInt needConvert; + + p->bufs[BCJ2_STREAM_MAIN] = dest + 1; + p->ip += (UInt32)num + 1; + src++; + + needConvert = False; + + if ((SizeT)(p->srcLim - src) >= 4) + { + UInt32 relatVal = GetUi32(src); + if ((p->fileSize == 0 || (UInt32)(p->ip + 4 + relatVal - p->fileIp) < p->fileSize) + && ((relatVal + p->relatLimit) >> 1) < p->relatLimit) + needConvert = True; + } + + { + UInt32 bound; + unsigned ttt; + Byte b = src[-1]; + CProb *prob = p->probs + (unsigned)(b == 0xE8 ? 2 + (unsigned)context : (b == 0xE9 ? 1 : 0)); + + ttt = *prob; + bound = (p->range >> kNumModelBits) * ttt; + + if (!needConvert) + { + p->range = bound; + *prob = (CProb)(ttt + ((kBitModelTotal - ttt) >> kNumMoveBits)); + p->src = src; + p->prevByte = b; + continue; + } + + p->low += bound; + p->range -= bound; + *prob = (CProb)(ttt - (ttt >> kNumMoveBits)); + + { + UInt32 relatVal = GetUi32(src); + UInt32 absVal; + p->ip += 4; + absVal = p->ip + relatVal; + p->prevByte = src[3]; + src += 4; + p->src = src; + { + unsigned cj = (b == 0xE8) ? BCJ2_STREAM_CALL : BCJ2_STREAM_JUMP; + Byte *cur = p->bufs[cj]; + if (cur == p->lims[cj]) + { + p->state = cj; + p->tempTarget = absVal; + return; + } + SetBe32(cur, absVal); + p->bufs[cj] = cur + 4; + } + } + } + } + } + } + } + + if (p->finishMode != BCJ2_ENC_FINISH_MODE_END_STREAM) + return; + + for (; p->flushPos < 5; p->flushPos++) + if (RangeEnc_ShiftLow(p)) + return; + p->state = BCJ2_ENC_STATE_OK; +} + + +void Bcj2Enc_Encode(CBcj2Enc *p) +{ + PRF(printf("\n")); + PRF(printf("---- ip = %8d tempPos = %8d src = %8d\n", p->ip, p->tempPos, p->srcLim - p->src)); + + if (p->tempPos != 0) + { + unsigned extra = 0; + + for (;;) + { + const Byte *src = p->src; + const Byte *srcLim = p->srcLim; + EBcj2Enc_FinishMode finishMode = p->finishMode; + + p->src = p->temp; + p->srcLim = p->temp + p->tempPos; + if (src != srcLim) + p->finishMode = BCJ2_ENC_FINISH_MODE_CONTINUE; + + PRF(printf(" ip = %8d tempPos = %8d src = %8d\n", p->ip, p->tempPos, p->srcLim - p->src)); + + Bcj2Enc_Encode_2(p); + + { + unsigned num = (unsigned)(p->src - p->temp); + unsigned tempPos = p->tempPos - num; + unsigned i; + p->tempPos = tempPos; + for (i = 0; i < tempPos; i++) + p->temp[i] = p->temp[(size_t)i + num]; + + p->src = src; + p->srcLim = srcLim; + p->finishMode = finishMode; + + if (p->state != BCJ2_ENC_STATE_ORIG || src == srcLim) + return; + + if (extra >= tempPos) + { + p->src = src - tempPos; + p->tempPos = 0; + break; + } + + p->temp[tempPos] = src[0]; + p->tempPos = tempPos + 1; + p->src = src + 1; + extra++; + } + } + } + + PRF(printf("++++ ip = %8d tempPos = %8d src = %8d\n", p->ip, p->tempPos, p->srcLim - p->src)); + + Bcj2Enc_Encode_2(p); + + if (p->state == BCJ2_ENC_STATE_ORIG) + { + const Byte *src = p->src; + unsigned rem = (unsigned)(p->srcLim - src); + unsigned i; + for (i = 0; i < rem; i++) + p->temp[i] = src[i]; + p->tempPos = rem; + p->src = src + rem; + } +} diff --git a/deps/LZMA-SDK/C/Bra.c b/deps/LZMA-SDK/C/Bra.c index cdefa4d2e..3b854d9ca 100644 --- a/deps/LZMA-SDK/C/Bra.c +++ b/deps/LZMA-SDK/C/Bra.c @@ -1,230 +1,230 @@ -/* Bra.c -- Converters for RISC code -2021-02-09 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "CpuArch.h" -#include "Bra.h" - -SizeT ARM_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) -{ - Byte *p; - const Byte *lim; - size &= ~(size_t)3; - ip += 4; - p = data; - lim = data + size; - - if (encoding) - - for (;;) - { - for (;;) - { - if (p >= lim) - return (SizeT)(p - data); - p += 4; - if (p[-1] == 0xEB) - break; - } - { - UInt32 v = GetUi32(p - 4); - v <<= 2; - v += ip + (UInt32)(p - data); - v >>= 2; - v &= 0x00FFFFFF; - v |= 0xEB000000; - SetUi32(p - 4, v); - } - } - - for (;;) - { - for (;;) - { - if (p >= lim) - return (SizeT)(p - data); - p += 4; - if (p[-1] == 0xEB) - break; - } - { - UInt32 v = GetUi32(p - 4); - v <<= 2; - v -= ip + (UInt32)(p - data); - v >>= 2; - v &= 0x00FFFFFF; - v |= 0xEB000000; - SetUi32(p - 4, v); - } - } -} - - -SizeT ARMT_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) -{ - Byte *p; - const Byte *lim; - size &= ~(size_t)1; - p = data; - lim = data + size - 4; - - if (encoding) - - for (;;) - { - UInt32 b1; - for (;;) - { - UInt32 b3; - if (p > lim) - return (SizeT)(p - data); - b1 = p[1]; - b3 = p[3]; - p += 2; - b1 ^= 8; - if ((b3 & b1) >= 0xF8) - break; - } - { - UInt32 v = - ((UInt32)b1 << 19) - + (((UInt32)p[1] & 0x7) << 8) - + (((UInt32)p[-2] << 11)) - + (p[0]); - - p += 2; - { - UInt32 cur = (ip + (UInt32)(p - data)) >> 1; - v += cur; - } - - p[-4] = (Byte)(v >> 11); - p[-3] = (Byte)(0xF0 | ((v >> 19) & 0x7)); - p[-2] = (Byte)v; - p[-1] = (Byte)(0xF8 | (v >> 8)); - } - } - - for (;;) - { - UInt32 b1; - for (;;) - { - UInt32 b3; - if (p > lim) - return (SizeT)(p - data); - b1 = p[1]; - b3 = p[3]; - p += 2; - b1 ^= 8; - if ((b3 & b1) >= 0xF8) - break; - } - { - UInt32 v = - ((UInt32)b1 << 19) - + (((UInt32)p[1] & 0x7) << 8) - + (((UInt32)p[-2] << 11)) - + (p[0]); - - p += 2; - { - UInt32 cur = (ip + (UInt32)(p - data)) >> 1; - v -= cur; - } - - /* - SetUi16(p - 4, (UInt16)(((v >> 11) & 0x7FF) | 0xF000)); - SetUi16(p - 2, (UInt16)(v | 0xF800)); - */ - - p[-4] = (Byte)(v >> 11); - p[-3] = (Byte)(0xF0 | ((v >> 19) & 0x7)); - p[-2] = (Byte)v; - p[-1] = (Byte)(0xF8 | (v >> 8)); - } - } -} - - -SizeT PPC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) -{ - Byte *p; - const Byte *lim; - size &= ~(size_t)3; - ip -= 4; - p = data; - lim = data + size; - - for (;;) - { - for (;;) - { - if (p >= lim) - return (SizeT)(p - data); - p += 4; - /* if ((v & 0xFC000003) == 0x48000001) */ - if ((p[-4] & 0xFC) == 0x48 && (p[-1] & 3) == 1) - break; - } - { - UInt32 v = GetBe32(p - 4); - if (encoding) - v += ip + (UInt32)(p - data); - else - v -= ip + (UInt32)(p - data); - v &= 0x03FFFFFF; - v |= 0x48000000; - SetBe32(p - 4, v); - } - } -} - - -SizeT SPARC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) -{ - Byte *p; - const Byte *lim; - size &= ~(size_t)3; - ip -= 4; - p = data; - lim = data + size; - - for (;;) - { - for (;;) - { - if (p >= lim) - return (SizeT)(p - data); - /* - v = GetBe32(p); - p += 4; - m = v + ((UInt32)5 << 29); - m ^= (UInt32)7 << 29; - m += (UInt32)1 << 22; - if ((m & ((UInt32)0x1FF << 23)) == 0) - break; - */ - p += 4; - if ((p[-4] == 0x40 && (p[-3] & 0xC0) == 0) || - (p[-4] == 0x7F && (p[-3] >= 0xC0))) - break; - } - { - UInt32 v = GetBe32(p - 4); - v <<= 2; - if (encoding) - v += ip + (UInt32)(p - data); - else - v -= ip + (UInt32)(p - data); - - v &= 0x01FFFFFF; - v -= (UInt32)1 << 24; - v ^= 0xFF000000; - v >>= 2; - v |= 0x40000000; - SetBe32(p - 4, v); - } - } -} +/* Bra.c -- Converters for RISC code +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" +#include "Bra.h" + +SizeT ARM_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + Byte *p; + const Byte *lim; + size &= ~(size_t)3; + ip += 4; + p = data; + lim = data + size; + + if (encoding) + + for (;;) + { + for (;;) + { + if (p >= lim) + return (SizeT)(p - data); + p += 4; + if (p[-1] == 0xEB) + break; + } + { + UInt32 v = GetUi32(p - 4); + v <<= 2; + v += ip + (UInt32)(p - data); + v >>= 2; + v &= 0x00FFFFFF; + v |= 0xEB000000; + SetUi32(p - 4, v); + } + } + + for (;;) + { + for (;;) + { + if (p >= lim) + return (SizeT)(p - data); + p += 4; + if (p[-1] == 0xEB) + break; + } + { + UInt32 v = GetUi32(p - 4); + v <<= 2; + v -= ip + (UInt32)(p - data); + v >>= 2; + v &= 0x00FFFFFF; + v |= 0xEB000000; + SetUi32(p - 4, v); + } + } +} + + +SizeT ARMT_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + Byte *p; + const Byte *lim; + size &= ~(size_t)1; + p = data; + lim = data + size - 4; + + if (encoding) + + for (;;) + { + UInt32 b1; + for (;;) + { + UInt32 b3; + if (p > lim) + return (SizeT)(p - data); + b1 = p[1]; + b3 = p[3]; + p += 2; + b1 ^= 8; + if ((b3 & b1) >= 0xF8) + break; + } + { + UInt32 v = + ((UInt32)b1 << 19) + + (((UInt32)p[1] & 0x7) << 8) + + (((UInt32)p[-2] << 11)) + + (p[0]); + + p += 2; + { + UInt32 cur = (ip + (UInt32)(p - data)) >> 1; + v += cur; + } + + p[-4] = (Byte)(v >> 11); + p[-3] = (Byte)(0xF0 | ((v >> 19) & 0x7)); + p[-2] = (Byte)v; + p[-1] = (Byte)(0xF8 | (v >> 8)); + } + } + + for (;;) + { + UInt32 b1; + for (;;) + { + UInt32 b3; + if (p > lim) + return (SizeT)(p - data); + b1 = p[1]; + b3 = p[3]; + p += 2; + b1 ^= 8; + if ((b3 & b1) >= 0xF8) + break; + } + { + UInt32 v = + ((UInt32)b1 << 19) + + (((UInt32)p[1] & 0x7) << 8) + + (((UInt32)p[-2] << 11)) + + (p[0]); + + p += 2; + { + UInt32 cur = (ip + (UInt32)(p - data)) >> 1; + v -= cur; + } + + /* + SetUi16(p - 4, (UInt16)(((v >> 11) & 0x7FF) | 0xF000)); + SetUi16(p - 2, (UInt16)(v | 0xF800)); + */ + + p[-4] = (Byte)(v >> 11); + p[-3] = (Byte)(0xF0 | ((v >> 19) & 0x7)); + p[-2] = (Byte)v; + p[-1] = (Byte)(0xF8 | (v >> 8)); + } + } +} + + +SizeT PPC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + Byte *p; + const Byte *lim; + size &= ~(size_t)3; + ip -= 4; + p = data; + lim = data + size; + + for (;;) + { + for (;;) + { + if (p >= lim) + return (SizeT)(p - data); + p += 4; + /* if ((v & 0xFC000003) == 0x48000001) */ + if ((p[-4] & 0xFC) == 0x48 && (p[-1] & 3) == 1) + break; + } + { + UInt32 v = GetBe32(p - 4); + if (encoding) + v += ip + (UInt32)(p - data); + else + v -= ip + (UInt32)(p - data); + v &= 0x03FFFFFF; + v |= 0x48000000; + SetBe32(p - 4, v); + } + } +} + + +SizeT SPARC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + Byte *p; + const Byte *lim; + size &= ~(size_t)3; + ip -= 4; + p = data; + lim = data + size; + + for (;;) + { + for (;;) + { + if (p >= lim) + return (SizeT)(p - data); + /* + v = GetBe32(p); + p += 4; + m = v + ((UInt32)5 << 29); + m ^= (UInt32)7 << 29; + m += (UInt32)1 << 22; + if ((m & ((UInt32)0x1FF << 23)) == 0) + break; + */ + p += 4; + if ((p[-4] == 0x40 && (p[-3] & 0xC0) == 0) || + (p[-4] == 0x7F && (p[-3] >= 0xC0))) + break; + } + { + UInt32 v = GetBe32(p - 4); + v <<= 2; + if (encoding) + v += ip + (UInt32)(p - data); + else + v -= ip + (UInt32)(p - data); + + v &= 0x01FFFFFF; + v -= (UInt32)1 << 24; + v ^= 0xFF000000; + v >>= 2; + v |= 0x40000000; + SetBe32(p - 4, v); + } + } +} diff --git a/deps/LZMA-SDK/C/Bra.h b/deps/LZMA-SDK/C/Bra.h index aba8dce14..855e37a6b 100644 --- a/deps/LZMA-SDK/C/Bra.h +++ b/deps/LZMA-SDK/C/Bra.h @@ -1,64 +1,64 @@ -/* Bra.h -- Branch converters for executables -2013-01-18 : Igor Pavlov : Public domain */ - -#ifndef __BRA_H -#define __BRA_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -/* -These functions convert relative addresses to absolute addresses -in CALL instructions to increase the compression ratio. - - In: - data - data buffer - size - size of data - ip - current virtual Instruction Pinter (IP) value - state - state variable for x86 converter - encoding - 0 (for decoding), 1 (for encoding) - - Out: - state - state variable for x86 converter - - Returns: - The number of processed bytes. If you call these functions with multiple calls, - you must start next call with first byte after block of processed bytes. - - Type Endian Alignment LookAhead - - x86 little 1 4 - ARMT little 2 2 - ARM little 4 0 - PPC big 4 0 - SPARC big 4 0 - IA64 little 16 0 - - size must be >= Alignment + LookAhead, if it's not last block. - If (size < Alignment + LookAhead), converter returns 0. - - Example: - - UInt32 ip = 0; - for () - { - ; size must be >= Alignment + LookAhead, if it's not last block - SizeT processed = Convert(data, size, ip, 1); - data += processed; - size -= processed; - ip += processed; - } -*/ - -#define x86_Convert_Init(state) { state = 0; } -SizeT x86_Convert(Byte *data, SizeT size, UInt32 ip, UInt32 *state, int encoding); -SizeT ARM_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); -SizeT ARMT_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); -SizeT PPC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); -SizeT SPARC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); -SizeT IA64_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); - -EXTERN_C_END - -#endif +/* Bra.h -- Branch converters for executables +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __BRA_H +#define __BRA_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +/* +These functions convert relative addresses to absolute addresses +in CALL instructions to increase the compression ratio. + + In: + data - data buffer + size - size of data + ip - current virtual Instruction Pinter (IP) value + state - state variable for x86 converter + encoding - 0 (for decoding), 1 (for encoding) + + Out: + state - state variable for x86 converter + + Returns: + The number of processed bytes. If you call these functions with multiple calls, + you must start next call with first byte after block of processed bytes. + + Type Endian Alignment LookAhead + + x86 little 1 4 + ARMT little 2 2 + ARM little 4 0 + PPC big 4 0 + SPARC big 4 0 + IA64 little 16 0 + + size must be >= Alignment + LookAhead, if it's not last block. + If (size < Alignment + LookAhead), converter returns 0. + + Example: + + UInt32 ip = 0; + for () + { + ; size must be >= Alignment + LookAhead, if it's not last block + SizeT processed = Convert(data, size, ip, 1); + data += processed; + size -= processed; + ip += processed; + } +*/ + +#define x86_Convert_Init(state) { state = 0; } +SizeT x86_Convert(Byte *data, SizeT size, UInt32 ip, UInt32 *state, int encoding); +SizeT ARM_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); +SizeT ARMT_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); +SizeT PPC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); +SizeT SPARC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); +SizeT IA64_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/Bra86.c b/deps/LZMA-SDK/C/Bra86.c index d857dac67..10a0fbd16 100644 --- a/deps/LZMA-SDK/C/Bra86.c +++ b/deps/LZMA-SDK/C/Bra86.c @@ -1,82 +1,82 @@ -/* Bra86.c -- Converter for x86 code (BCJ) -2021-02-09 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "Bra.h" - -#define Test86MSByte(b) ((((b) + 1) & 0xFE) == 0) - -SizeT x86_Convert(Byte *data, SizeT size, UInt32 ip, UInt32 *state, int encoding) -{ - SizeT pos = 0; - UInt32 mask = *state & 7; - if (size < 5) - return 0; - size -= 4; - ip += 5; - - for (;;) - { - Byte *p = data + pos; - const Byte *limit = data + size; - for (; p < limit; p++) - if ((*p & 0xFE) == 0xE8) - break; - - { - SizeT d = (SizeT)(p - data) - pos; - pos = (SizeT)(p - data); - if (p >= limit) - { - *state = (d > 2 ? 0 : mask >> (unsigned)d); - return pos; - } - if (d > 2) - mask = 0; - else - { - mask >>= (unsigned)d; - if (mask != 0 && (mask > 4 || mask == 3 || Test86MSByte(p[(size_t)(mask >> 1) + 1]))) - { - mask = (mask >> 1) | 4; - pos++; - continue; - } - } - } - - if (Test86MSByte(p[4])) - { - UInt32 v = ((UInt32)p[4] << 24) | ((UInt32)p[3] << 16) | ((UInt32)p[2] << 8) | ((UInt32)p[1]); - UInt32 cur = ip + (UInt32)pos; - pos += 5; - if (encoding) - v += cur; - else - v -= cur; - if (mask != 0) - { - unsigned sh = (mask & 6) << 2; - if (Test86MSByte((Byte)(v >> sh))) - { - v ^= (((UInt32)0x100 << sh) - 1); - if (encoding) - v += cur; - else - v -= cur; - } - mask = 0; - } - p[1] = (Byte)v; - p[2] = (Byte)(v >> 8); - p[3] = (Byte)(v >> 16); - p[4] = (Byte)(0 - ((v >> 24) & 1)); - } - else - { - mask = (mask >> 1) | 4; - pos++; - } - } -} +/* Bra86.c -- Converter for x86 code (BCJ) +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "Bra.h" + +#define Test86MSByte(b) ((((b) + 1) & 0xFE) == 0) + +SizeT x86_Convert(Byte *data, SizeT size, UInt32 ip, UInt32 *state, int encoding) +{ + SizeT pos = 0; + UInt32 mask = *state & 7; + if (size < 5) + return 0; + size -= 4; + ip += 5; + + for (;;) + { + Byte *p = data + pos; + const Byte *limit = data + size; + for (; p < limit; p++) + if ((*p & 0xFE) == 0xE8) + break; + + { + SizeT d = (SizeT)(p - data) - pos; + pos = (SizeT)(p - data); + if (p >= limit) + { + *state = (d > 2 ? 0 : mask >> (unsigned)d); + return pos; + } + if (d > 2) + mask = 0; + else + { + mask >>= (unsigned)d; + if (mask != 0 && (mask > 4 || mask == 3 || Test86MSByte(p[(size_t)(mask >> 1) + 1]))) + { + mask = (mask >> 1) | 4; + pos++; + continue; + } + } + } + + if (Test86MSByte(p[4])) + { + UInt32 v = ((UInt32)p[4] << 24) | ((UInt32)p[3] << 16) | ((UInt32)p[2] << 8) | ((UInt32)p[1]); + UInt32 cur = ip + (UInt32)pos; + pos += 5; + if (encoding) + v += cur; + else + v -= cur; + if (mask != 0) + { + unsigned sh = (mask & 6) << 2; + if (Test86MSByte((Byte)(v >> sh))) + { + v ^= (((UInt32)0x100 << sh) - 1); + if (encoding) + v += cur; + else + v -= cur; + } + mask = 0; + } + p[1] = (Byte)v; + p[2] = (Byte)(v >> 8); + p[3] = (Byte)(v >> 16); + p[4] = (Byte)(0 - ((v >> 24) & 1)); + } + else + { + mask = (mask >> 1) | 4; + pos++; + } + } +} diff --git a/deps/LZMA-SDK/C/BraIA64.c b/deps/LZMA-SDK/C/BraIA64.c index 2656907a0..d1dbc62c5 100644 --- a/deps/LZMA-SDK/C/BraIA64.c +++ b/deps/LZMA-SDK/C/BraIA64.c @@ -1,53 +1,53 @@ -/* BraIA64.c -- Converter for IA-64 code -2017-01-26 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "CpuArch.h" -#include "Bra.h" - -SizeT IA64_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) -{ - SizeT i; - if (size < 16) - return 0; - size -= 16; - i = 0; - do - { - unsigned m = ((UInt32)0x334B0000 >> (data[i] & 0x1E)) & 3; - if (m) - { - m++; - do - { - Byte *p = data + (i + (size_t)m * 5 - 8); - if (((p[3] >> m) & 15) == 5 - && (((p[-1] | ((UInt32)p[0] << 8)) >> m) & 0x70) == 0) - { - unsigned raw = GetUi32(p); - unsigned v = raw >> m; - v = (v & 0xFFFFF) | ((v & (1 << 23)) >> 3); - - v <<= 4; - if (encoding) - v += ip + (UInt32)i; - else - v -= ip + (UInt32)i; - v >>= 4; - - v &= 0x1FFFFF; - v += 0x700000; - v &= 0x8FFFFF; - raw &= ~((UInt32)0x8FFFFF << m); - raw |= (v << m); - SetUi32(p, raw); - } - } - while (++m <= 4); - } - i += 16; - } - while (i <= size); - return i; -} +/* BraIA64.c -- Converter for IA-64 code +2017-01-26 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" +#include "Bra.h" + +SizeT IA64_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + SizeT i; + if (size < 16) + return 0; + size -= 16; + i = 0; + do + { + unsigned m = ((UInt32)0x334B0000 >> (data[i] & 0x1E)) & 3; + if (m) + { + m++; + do + { + Byte *p = data + (i + (size_t)m * 5 - 8); + if (((p[3] >> m) & 15) == 5 + && (((p[-1] | ((UInt32)p[0] << 8)) >> m) & 0x70) == 0) + { + unsigned raw = GetUi32(p); + unsigned v = raw >> m; + v = (v & 0xFFFFF) | ((v & (1 << 23)) >> 3); + + v <<= 4; + if (encoding) + v += ip + (UInt32)i; + else + v -= ip + (UInt32)i; + v >>= 4; + + v &= 0x1FFFFF; + v += 0x700000; + v &= 0x8FFFFF; + raw &= ~((UInt32)0x8FFFFF << m); + raw |= (v << m); + SetUi32(p, raw); + } + } + while (++m <= 4); + } + i += 16; + } + while (i <= size); + return i; +} diff --git a/deps/LZMA-SDK/C/Compiler.h b/deps/LZMA-SDK/C/Compiler.h index eba374298..a9816fa5a 100644 --- a/deps/LZMA-SDK/C/Compiler.h +++ b/deps/LZMA-SDK/C/Compiler.h @@ -1,43 +1,43 @@ -/* Compiler.h -2021-01-05 : Igor Pavlov : Public domain */ - -#ifndef __7Z_COMPILER_H -#define __7Z_COMPILER_H - - #ifdef __clang__ - #pragma clang diagnostic ignored "-Wunused-private-field" - #endif - -#ifdef _MSC_VER - - #ifdef UNDER_CE - #define RPC_NO_WINDOWS_H - /* #pragma warning(disable : 4115) // '_RPC_ASYNC_STATE' : named type definition in parentheses */ - #pragma warning(disable : 4201) // nonstandard extension used : nameless struct/union - #pragma warning(disable : 4214) // nonstandard extension used : bit field types other than int - #endif - - #if _MSC_VER >= 1300 - #pragma warning(disable : 4996) // This function or variable may be unsafe - #else - #pragma warning(disable : 4511) // copy constructor could not be generated - #pragma warning(disable : 4512) // assignment operator could not be generated - #pragma warning(disable : 4514) // unreferenced inline function has been removed - #pragma warning(disable : 4702) // unreachable code - #pragma warning(disable : 4710) // not inlined - #pragma warning(disable : 4714) // function marked as __forceinline not inlined - #pragma warning(disable : 4786) // identifier was truncated to '255' characters in the debug information - #endif - - #ifdef __clang__ - #pragma clang diagnostic ignored "-Wdeprecated-declarations" - #pragma clang diagnostic ignored "-Wmicrosoft-exception-spec" - // #pragma clang diagnostic ignored "-Wreserved-id-macro" - #endif - -#endif - -#define UNUSED_VAR(x) (void)x; -/* #define UNUSED_VAR(x) x=x; */ - -#endif +/* Compiler.h +2021-01-05 : Igor Pavlov : Public domain */ + +#ifndef __7Z_COMPILER_H +#define __7Z_COMPILER_H + + #ifdef __clang__ + #pragma clang diagnostic ignored "-Wunused-private-field" + #endif + +#ifdef _MSC_VER + + #ifdef UNDER_CE + #define RPC_NO_WINDOWS_H + /* #pragma warning(disable : 4115) // '_RPC_ASYNC_STATE' : named type definition in parentheses */ + #pragma warning(disable : 4201) // nonstandard extension used : nameless struct/union + #pragma warning(disable : 4214) // nonstandard extension used : bit field types other than int + #endif + + #if _MSC_VER >= 1300 + #pragma warning(disable : 4996) // This function or variable may be unsafe + #else + #pragma warning(disable : 4511) // copy constructor could not be generated + #pragma warning(disable : 4512) // assignment operator could not be generated + #pragma warning(disable : 4514) // unreferenced inline function has been removed + #pragma warning(disable : 4702) // unreachable code + #pragma warning(disable : 4710) // not inlined + #pragma warning(disable : 4714) // function marked as __forceinline not inlined + #pragma warning(disable : 4786) // identifier was truncated to '255' characters in the debug information + #endif + + #ifdef __clang__ + #pragma clang diagnostic ignored "-Wdeprecated-declarations" + #pragma clang diagnostic ignored "-Wmicrosoft-exception-spec" + // #pragma clang diagnostic ignored "-Wreserved-id-macro" + #endif + +#endif + +#define UNUSED_VAR(x) (void)x; +/* #define UNUSED_VAR(x) x=x; */ + +#endif diff --git a/deps/LZMA-SDK/C/CpuArch.c b/deps/LZMA-SDK/C/CpuArch.c index 36e0be0b1..110ced6d6 100644 --- a/deps/LZMA-SDK/C/CpuArch.c +++ b/deps/LZMA-SDK/C/CpuArch.c @@ -1,478 +1,478 @@ -/* CpuArch.c -- CPU specific code -2021-07-13 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "CpuArch.h" - -#ifdef MY_CPU_X86_OR_AMD64 - -#if (defined(_MSC_VER) && !defined(MY_CPU_AMD64)) || defined(__GNUC__) -#define USE_ASM -#endif - -#if !defined(USE_ASM) && _MSC_VER >= 1500 -#include -#endif - -#if defined(USE_ASM) && !defined(MY_CPU_AMD64) -static UInt32 CheckFlag(UInt32 flag) -{ - #ifdef _MSC_VER - __asm pushfd; - __asm pop EAX; - __asm mov EDX, EAX; - __asm xor EAX, flag; - __asm push EAX; - __asm popfd; - __asm pushfd; - __asm pop EAX; - __asm xor EAX, EDX; - __asm push EDX; - __asm popfd; - __asm and flag, EAX; - #else - __asm__ __volatile__ ( - "pushf\n\t" - "pop %%EAX\n\t" - "movl %%EAX,%%EDX\n\t" - "xorl %0,%%EAX\n\t" - "push %%EAX\n\t" - "popf\n\t" - "pushf\n\t" - "pop %%EAX\n\t" - "xorl %%EDX,%%EAX\n\t" - "push %%EDX\n\t" - "popf\n\t" - "andl %%EAX, %0\n\t": - "=c" (flag) : "c" (flag) : - "%eax", "%edx"); - #endif - return flag; -} -#define CHECK_CPUID_IS_SUPPORTED if (CheckFlag(1 << 18) == 0 || CheckFlag(1 << 21) == 0) return False; -#else -#define CHECK_CPUID_IS_SUPPORTED -#endif - -#ifndef USE_ASM - #ifdef _MSC_VER - #if _MSC_VER >= 1600 - #define MY__cpuidex __cpuidex - #else - -/* - __cpuid (function == 4) requires subfunction number in ECX. - MSDN: The __cpuid intrinsic clears the ECX register before calling the cpuid instruction. - __cpuid() in new MSVC clears ECX. - __cpuid() in old MSVC (14.00) doesn't clear ECX - We still can use __cpuid for low (function) values that don't require ECX, - but __cpuid() in old MSVC will be incorrect for some function values: (function == 4). - So here we use the hack for old MSVC to send (subFunction) in ECX register to cpuid instruction, - where ECX value is first parameter for FAST_CALL / NO_INLINE function, - So the caller of MY__cpuidex_HACK() sets ECX as subFunction, and - old MSVC for __cpuid() doesn't change ECX and cpuid instruction gets (subFunction) value. - - DON'T remove MY_NO_INLINE and MY_FAST_CALL for MY__cpuidex_HACK() !!! -*/ - -static -MY_NO_INLINE -void MY_FAST_CALL MY__cpuidex_HACK(UInt32 subFunction, int *CPUInfo, UInt32 function) -{ - UNUSED_VAR(subFunction); - __cpuid(CPUInfo, function); -} - - #define MY__cpuidex(info, func, func2) MY__cpuidex_HACK(func2, info, func) - #pragma message("======== MY__cpuidex_HACK WAS USED ========") - #endif - #else - #define MY__cpuidex(info, func, func2) __cpuid(info, func) - #pragma message("======== (INCORRECT ?) cpuid WAS USED ========") - #endif -#endif - - - - -void MyCPUID(UInt32 function, UInt32 *a, UInt32 *b, UInt32 *c, UInt32 *d) -{ - #ifdef USE_ASM - - #ifdef _MSC_VER - - UInt32 a2, b2, c2, d2; - __asm xor EBX, EBX; - __asm xor ECX, ECX; - __asm xor EDX, EDX; - __asm mov EAX, function; - __asm cpuid; - __asm mov a2, EAX; - __asm mov b2, EBX; - __asm mov c2, ECX; - __asm mov d2, EDX; - - *a = a2; - *b = b2; - *c = c2; - *d = d2; - - #else - - __asm__ __volatile__ ( - #if defined(MY_CPU_AMD64) && defined(__PIC__) - "mov %%rbx, %%rdi;" - "cpuid;" - "xchg %%rbx, %%rdi;" - : "=a" (*a) , - "=D" (*b) , - #elif defined(MY_CPU_X86) && defined(__PIC__) - "mov %%ebx, %%edi;" - "cpuid;" - "xchgl %%ebx, %%edi;" - : "=a" (*a) , - "=D" (*b) , - #else - "cpuid" - : "=a" (*a) , - "=b" (*b) , - #endif - "=c" (*c) , - "=d" (*d) - : "0" (function), "c"(0) ) ; - - #endif - - #else - - int CPUInfo[4]; - - MY__cpuidex(CPUInfo, (int)function, 0); - - *a = (UInt32)CPUInfo[0]; - *b = (UInt32)CPUInfo[1]; - *c = (UInt32)CPUInfo[2]; - *d = (UInt32)CPUInfo[3]; - - #endif -} - -BoolInt x86cpuid_CheckAndRead(Cx86cpuid *p) -{ - CHECK_CPUID_IS_SUPPORTED - MyCPUID(0, &p->maxFunc, &p->vendor[0], &p->vendor[2], &p->vendor[1]); - MyCPUID(1, &p->ver, &p->b, &p->c, &p->d); - return True; -} - -static const UInt32 kVendors[][3] = -{ - { 0x756E6547, 0x49656E69, 0x6C65746E}, - { 0x68747541, 0x69746E65, 0x444D4163}, - { 0x746E6543, 0x48727561, 0x736C7561} -}; - -int x86cpuid_GetFirm(const Cx86cpuid *p) -{ - unsigned i; - for (i = 0; i < sizeof(kVendors) / sizeof(kVendors[i]); i++) - { - const UInt32 *v = kVendors[i]; - if (v[0] == p->vendor[0] && - v[1] == p->vendor[1] && - v[2] == p->vendor[2]) - return (int)i; - } - return -1; -} - -BoolInt CPU_Is_InOrder() -{ - Cx86cpuid p; - int firm; - UInt32 family, model; - if (!x86cpuid_CheckAndRead(&p)) - return True; - - family = x86cpuid_GetFamily(p.ver); - model = x86cpuid_GetModel(p.ver); - - firm = x86cpuid_GetFirm(&p); - - switch (firm) - { - case CPU_FIRM_INTEL: return (family < 6 || (family == 6 && ( - /* In-Order Atom CPU */ - model == 0x1C /* 45 nm, N4xx, D4xx, N5xx, D5xx, 230, 330 */ - || model == 0x26 /* 45 nm, Z6xx */ - || model == 0x27 /* 32 nm, Z2460 */ - || model == 0x35 /* 32 nm, Z2760 */ - || model == 0x36 /* 32 nm, N2xxx, D2xxx */ - ))); - case CPU_FIRM_AMD: return (family < 5 || (family == 5 && (model < 6 || model == 0xA))); - case CPU_FIRM_VIA: return (family < 6 || (family == 6 && model < 0xF)); - } - return True; -} - -#if !defined(MY_CPU_AMD64) && defined(_WIN32) -#include -static BoolInt CPU_Sys_Is_SSE_Supported() -{ - OSVERSIONINFO vi; - vi.dwOSVersionInfoSize = sizeof(vi); - if (!GetVersionEx(&vi)) - return False; - return (vi.dwMajorVersion >= 5); -} -#define CHECK_SYS_SSE_SUPPORT if (!CPU_Sys_Is_SSE_Supported()) return False; -#else -#define CHECK_SYS_SSE_SUPPORT -#endif - - -static UInt32 X86_CPUID_ECX_Get_Flags() -{ - Cx86cpuid p; - CHECK_SYS_SSE_SUPPORT - if (!x86cpuid_CheckAndRead(&p)) - return 0; - return p.c; -} - -BoolInt CPU_IsSupported_AES() -{ - return (X86_CPUID_ECX_Get_Flags() >> 25) & 1; -} - -BoolInt CPU_IsSupported_SSSE3() -{ - return (X86_CPUID_ECX_Get_Flags() >> 9) & 1; -} - -BoolInt CPU_IsSupported_SSE41() -{ - return (X86_CPUID_ECX_Get_Flags() >> 19) & 1; -} - -BoolInt CPU_IsSupported_SHA() -{ - Cx86cpuid p; - CHECK_SYS_SSE_SUPPORT - if (!x86cpuid_CheckAndRead(&p)) - return False; - - if (p.maxFunc < 7) - return False; - { - UInt32 d[4] = { 0 }; - MyCPUID(7, &d[0], &d[1], &d[2], &d[3]); - return (d[1] >> 29) & 1; - } -} - -// #include - -#ifdef _WIN32 -#include -#endif - -BoolInt CPU_IsSupported_AVX2() -{ - Cx86cpuid p; - CHECK_SYS_SSE_SUPPORT - - #ifdef _WIN32 - #define MY__PF_XSAVE_ENABLED 17 - if (!IsProcessorFeaturePresent(MY__PF_XSAVE_ENABLED)) - return False; - #endif - - if (!x86cpuid_CheckAndRead(&p)) - return False; - if (p.maxFunc < 7) - return False; - { - UInt32 d[4] = { 0 }; - MyCPUID(7, &d[0], &d[1], &d[2], &d[3]); - // printf("\ncpuid(7): ebx=%8x ecx=%8x\n", d[1], d[2]); - return 1 - & (d[1] >> 5); // avx2 - } -} - -BoolInt CPU_IsSupported_VAES_AVX2() -{ - Cx86cpuid p; - CHECK_SYS_SSE_SUPPORT - - #ifdef _WIN32 - #define MY__PF_XSAVE_ENABLED 17 - if (!IsProcessorFeaturePresent(MY__PF_XSAVE_ENABLED)) - return False; - #endif - - if (!x86cpuid_CheckAndRead(&p)) - return False; - if (p.maxFunc < 7) - return False; - { - UInt32 d[4] = { 0 }; - MyCPUID(7, &d[0], &d[1], &d[2], &d[3]); - // printf("\ncpuid(7): ebx=%8x ecx=%8x\n", d[1], d[2]); - return 1 - & (d[1] >> 5) // avx2 - // & (d[1] >> 31) // avx512vl - & (d[2] >> 9); // vaes // VEX-256/EVEX - } -} - -BoolInt CPU_IsSupported_PageGB() -{ - Cx86cpuid cpuid; - if (!x86cpuid_CheckAndRead(&cpuid)) - return False; - { - UInt32 d[4] = { 0 }; - MyCPUID(0x80000000, &d[0], &d[1], &d[2], &d[3]); - if (d[0] < 0x80000001) - return False; - } - { - UInt32 d[4] = { 0 }; - MyCPUID(0x80000001, &d[0], &d[1], &d[2], &d[3]); - return (d[3] >> 26) & 1; - } -} - - -#elif defined(MY_CPU_ARM_OR_ARM64) - -#ifdef _WIN32 - -#include - -BoolInt CPU_IsSupported_CRC32() { return IsProcessorFeaturePresent(PF_ARM_V8_CRC32_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } -BoolInt CPU_IsSupported_CRYPTO() { return IsProcessorFeaturePresent(PF_ARM_V8_CRYPTO_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } -BoolInt CPU_IsSupported_NEON() { return IsProcessorFeaturePresent(PF_ARM_NEON_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } - -#else - -#if defined(__APPLE__) - -/* -#include -#include -static void Print_sysctlbyname(const char *name) -{ - size_t bufSize = 256; - char buf[256]; - int res = sysctlbyname(name, &buf, &bufSize, NULL, 0); - { - int i; - printf("\nres = %d : %s : '%s' : bufSize = %d, numeric", res, name, buf, (unsigned)bufSize); - for (i = 0; i < 20; i++) - printf(" %2x", (unsigned)(Byte)buf[i]); - - } -} -*/ - -static BoolInt My_sysctlbyname_Get_BoolInt(const char *name) -{ - UInt32 val = 0; - if (My_sysctlbyname_Get_UInt32(name, &val) == 0 && val == 1) - return 1; - return 0; -} - - /* - Print_sysctlbyname("hw.pagesize"); - Print_sysctlbyname("machdep.cpu.brand_string"); - */ - -BoolInt CPU_IsSupported_CRC32(void) -{ - return My_sysctlbyname_Get_BoolInt("hw.optional.armv8_crc32"); -} - -BoolInt CPU_IsSupported_NEON(void) -{ - return My_sysctlbyname_Get_BoolInt("hw.optional.neon"); -} - -#ifdef MY_CPU_ARM64 -#define APPLE_CRYPTO_SUPPORT_VAL 1 -#else -#define APPLE_CRYPTO_SUPPORT_VAL 0 -#endif - -BoolInt CPU_IsSupported_SHA1(void) { return APPLE_CRYPTO_SUPPORT_VAL; } -BoolInt CPU_IsSupported_SHA2(void) { return APPLE_CRYPTO_SUPPORT_VAL; } -BoolInt CPU_IsSupported_AES (void) { return APPLE_CRYPTO_SUPPORT_VAL; } - - -#else // __APPLE__ - -#include - -#define USE_HWCAP - -#ifdef USE_HWCAP - -#include - - #define MY_HWCAP_CHECK_FUNC_2(name1, name2) \ - BoolInt CPU_IsSupported_ ## name1() { return (getauxval(AT_HWCAP) & (HWCAP_ ## name2)) ? 1 : 0; } - -#ifdef MY_CPU_ARM64 - #define MY_HWCAP_CHECK_FUNC(name) \ - MY_HWCAP_CHECK_FUNC_2(name, name) - MY_HWCAP_CHECK_FUNC_2(NEON, ASIMD) -// MY_HWCAP_CHECK_FUNC (ASIMD) -#elif defined(MY_CPU_ARM) - #define MY_HWCAP_CHECK_FUNC(name) \ - BoolInt CPU_IsSupported_ ## name() { return (getauxval(AT_HWCAP2) & (HWCAP2_ ## name)) ? 1 : 0; } - MY_HWCAP_CHECK_FUNC_2(NEON, NEON) -#endif - -#else // USE_HWCAP - - #define MY_HWCAP_CHECK_FUNC(name) \ - BoolInt CPU_IsSupported_ ## name() { return 0; } - MY_HWCAP_CHECK_FUNC(NEON) - -#endif // USE_HWCAP - -MY_HWCAP_CHECK_FUNC (CRC32) -MY_HWCAP_CHECK_FUNC (SHA1) -MY_HWCAP_CHECK_FUNC (SHA2) -MY_HWCAP_CHECK_FUNC (AES) - -#endif // __APPLE__ -#endif // _WIN32 - -#endif // MY_CPU_ARM_OR_ARM64 - - - -#ifdef __APPLE__ - -#include - -int My_sysctlbyname_Get(const char *name, void *buf, size_t *bufSize) -{ - return sysctlbyname(name, buf, bufSize, NULL, 0); -} - -int My_sysctlbyname_Get_UInt32(const char *name, UInt32 *val) -{ - size_t bufSize = sizeof(*val); - int res = My_sysctlbyname_Get(name, val, &bufSize); - if (res == 0 && bufSize != sizeof(*val)) - return EFAULT; - return res; -} - -#endif +/* CpuArch.c -- CPU specific code +2021-07-13 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" + +#ifdef MY_CPU_X86_OR_AMD64 + +#if (defined(_MSC_VER) && !defined(MY_CPU_AMD64)) || defined(__GNUC__) +#define USE_ASM +#endif + +#if !defined(USE_ASM) && _MSC_VER >= 1500 +#include +#endif + +#if defined(USE_ASM) && !defined(MY_CPU_AMD64) +static UInt32 CheckFlag(UInt32 flag) +{ + #ifdef _MSC_VER + __asm pushfd; + __asm pop EAX; + __asm mov EDX, EAX; + __asm xor EAX, flag; + __asm push EAX; + __asm popfd; + __asm pushfd; + __asm pop EAX; + __asm xor EAX, EDX; + __asm push EDX; + __asm popfd; + __asm and flag, EAX; + #else + __asm__ __volatile__ ( + "pushf\n\t" + "pop %%EAX\n\t" + "movl %%EAX,%%EDX\n\t" + "xorl %0,%%EAX\n\t" + "push %%EAX\n\t" + "popf\n\t" + "pushf\n\t" + "pop %%EAX\n\t" + "xorl %%EDX,%%EAX\n\t" + "push %%EDX\n\t" + "popf\n\t" + "andl %%EAX, %0\n\t": + "=c" (flag) : "c" (flag) : + "%eax", "%edx"); + #endif + return flag; +} +#define CHECK_CPUID_IS_SUPPORTED if (CheckFlag(1 << 18) == 0 || CheckFlag(1 << 21) == 0) return False; +#else +#define CHECK_CPUID_IS_SUPPORTED +#endif + +#ifndef USE_ASM + #ifdef _MSC_VER + #if _MSC_VER >= 1600 + #define MY__cpuidex __cpuidex + #else + +/* + __cpuid (function == 4) requires subfunction number in ECX. + MSDN: The __cpuid intrinsic clears the ECX register before calling the cpuid instruction. + __cpuid() in new MSVC clears ECX. + __cpuid() in old MSVC (14.00) doesn't clear ECX + We still can use __cpuid for low (function) values that don't require ECX, + but __cpuid() in old MSVC will be incorrect for some function values: (function == 4). + So here we use the hack for old MSVC to send (subFunction) in ECX register to cpuid instruction, + where ECX value is first parameter for FAST_CALL / NO_INLINE function, + So the caller of MY__cpuidex_HACK() sets ECX as subFunction, and + old MSVC for __cpuid() doesn't change ECX and cpuid instruction gets (subFunction) value. + + DON'T remove MY_NO_INLINE and MY_FAST_CALL for MY__cpuidex_HACK() !!! +*/ + +static +MY_NO_INLINE +void MY_FAST_CALL MY__cpuidex_HACK(UInt32 subFunction, int *CPUInfo, UInt32 function) +{ + UNUSED_VAR(subFunction); + __cpuid(CPUInfo, function); +} + + #define MY__cpuidex(info, func, func2) MY__cpuidex_HACK(func2, info, func) + #pragma message("======== MY__cpuidex_HACK WAS USED ========") + #endif + #else + #define MY__cpuidex(info, func, func2) __cpuid(info, func) + #pragma message("======== (INCORRECT ?) cpuid WAS USED ========") + #endif +#endif + + + + +void MyCPUID(UInt32 function, UInt32 *a, UInt32 *b, UInt32 *c, UInt32 *d) +{ + #ifdef USE_ASM + + #ifdef _MSC_VER + + UInt32 a2, b2, c2, d2; + __asm xor EBX, EBX; + __asm xor ECX, ECX; + __asm xor EDX, EDX; + __asm mov EAX, function; + __asm cpuid; + __asm mov a2, EAX; + __asm mov b2, EBX; + __asm mov c2, ECX; + __asm mov d2, EDX; + + *a = a2; + *b = b2; + *c = c2; + *d = d2; + + #else + + __asm__ __volatile__ ( + #if defined(MY_CPU_AMD64) && defined(__PIC__) + "mov %%rbx, %%rdi;" + "cpuid;" + "xchg %%rbx, %%rdi;" + : "=a" (*a) , + "=D" (*b) , + #elif defined(MY_CPU_X86) && defined(__PIC__) + "mov %%ebx, %%edi;" + "cpuid;" + "xchgl %%ebx, %%edi;" + : "=a" (*a) , + "=D" (*b) , + #else + "cpuid" + : "=a" (*a) , + "=b" (*b) , + #endif + "=c" (*c) , + "=d" (*d) + : "0" (function), "c"(0) ) ; + + #endif + + #else + + int CPUInfo[4]; + + MY__cpuidex(CPUInfo, (int)function, 0); + + *a = (UInt32)CPUInfo[0]; + *b = (UInt32)CPUInfo[1]; + *c = (UInt32)CPUInfo[2]; + *d = (UInt32)CPUInfo[3]; + + #endif +} + +BoolInt x86cpuid_CheckAndRead(Cx86cpuid *p) +{ + CHECK_CPUID_IS_SUPPORTED + MyCPUID(0, &p->maxFunc, &p->vendor[0], &p->vendor[2], &p->vendor[1]); + MyCPUID(1, &p->ver, &p->b, &p->c, &p->d); + return True; +} + +static const UInt32 kVendors[][3] = +{ + { 0x756E6547, 0x49656E69, 0x6C65746E}, + { 0x68747541, 0x69746E65, 0x444D4163}, + { 0x746E6543, 0x48727561, 0x736C7561} +}; + +int x86cpuid_GetFirm(const Cx86cpuid *p) +{ + unsigned i; + for (i = 0; i < sizeof(kVendors) / sizeof(kVendors[i]); i++) + { + const UInt32 *v = kVendors[i]; + if (v[0] == p->vendor[0] && + v[1] == p->vendor[1] && + v[2] == p->vendor[2]) + return (int)i; + } + return -1; +} + +BoolInt CPU_Is_InOrder() +{ + Cx86cpuid p; + int firm; + UInt32 family, model; + if (!x86cpuid_CheckAndRead(&p)) + return True; + + family = x86cpuid_GetFamily(p.ver); + model = x86cpuid_GetModel(p.ver); + + firm = x86cpuid_GetFirm(&p); + + switch (firm) + { + case CPU_FIRM_INTEL: return (family < 6 || (family == 6 && ( + /* In-Order Atom CPU */ + model == 0x1C /* 45 nm, N4xx, D4xx, N5xx, D5xx, 230, 330 */ + || model == 0x26 /* 45 nm, Z6xx */ + || model == 0x27 /* 32 nm, Z2460 */ + || model == 0x35 /* 32 nm, Z2760 */ + || model == 0x36 /* 32 nm, N2xxx, D2xxx */ + ))); + case CPU_FIRM_AMD: return (family < 5 || (family == 5 && (model < 6 || model == 0xA))); + case CPU_FIRM_VIA: return (family < 6 || (family == 6 && model < 0xF)); + } + return True; +} + +#if !defined(MY_CPU_AMD64) && defined(_WIN32) +#include +static BoolInt CPU_Sys_Is_SSE_Supported() +{ + OSVERSIONINFO vi; + vi.dwOSVersionInfoSize = sizeof(vi); + if (!GetVersionEx(&vi)) + return False; + return (vi.dwMajorVersion >= 5); +} +#define CHECK_SYS_SSE_SUPPORT if (!CPU_Sys_Is_SSE_Supported()) return False; +#else +#define CHECK_SYS_SSE_SUPPORT +#endif + + +static UInt32 X86_CPUID_ECX_Get_Flags() +{ + Cx86cpuid p; + CHECK_SYS_SSE_SUPPORT + if (!x86cpuid_CheckAndRead(&p)) + return 0; + return p.c; +} + +BoolInt CPU_IsSupported_AES() +{ + return (X86_CPUID_ECX_Get_Flags() >> 25) & 1; +} + +BoolInt CPU_IsSupported_SSSE3() +{ + return (X86_CPUID_ECX_Get_Flags() >> 9) & 1; +} + +BoolInt CPU_IsSupported_SSE41() +{ + return (X86_CPUID_ECX_Get_Flags() >> 19) & 1; +} + +BoolInt CPU_IsSupported_SHA() +{ + Cx86cpuid p; + CHECK_SYS_SSE_SUPPORT + if (!x86cpuid_CheckAndRead(&p)) + return False; + + if (p.maxFunc < 7) + return False; + { + UInt32 d[4] = { 0 }; + MyCPUID(7, &d[0], &d[1], &d[2], &d[3]); + return (d[1] >> 29) & 1; + } +} + +// #include + +#ifdef _WIN32 +#include +#endif + +BoolInt CPU_IsSupported_AVX2() +{ + Cx86cpuid p; + CHECK_SYS_SSE_SUPPORT + + #ifdef _WIN32 + #define MY__PF_XSAVE_ENABLED 17 + if (!IsProcessorFeaturePresent(MY__PF_XSAVE_ENABLED)) + return False; + #endif + + if (!x86cpuid_CheckAndRead(&p)) + return False; + if (p.maxFunc < 7) + return False; + { + UInt32 d[4] = { 0 }; + MyCPUID(7, &d[0], &d[1], &d[2], &d[3]); + // printf("\ncpuid(7): ebx=%8x ecx=%8x\n", d[1], d[2]); + return 1 + & (d[1] >> 5); // avx2 + } +} + +BoolInt CPU_IsSupported_VAES_AVX2() +{ + Cx86cpuid p; + CHECK_SYS_SSE_SUPPORT + + #ifdef _WIN32 + #define MY__PF_XSAVE_ENABLED 17 + if (!IsProcessorFeaturePresent(MY__PF_XSAVE_ENABLED)) + return False; + #endif + + if (!x86cpuid_CheckAndRead(&p)) + return False; + if (p.maxFunc < 7) + return False; + { + UInt32 d[4] = { 0 }; + MyCPUID(7, &d[0], &d[1], &d[2], &d[3]); + // printf("\ncpuid(7): ebx=%8x ecx=%8x\n", d[1], d[2]); + return 1 + & (d[1] >> 5) // avx2 + // & (d[1] >> 31) // avx512vl + & (d[2] >> 9); // vaes // VEX-256/EVEX + } +} + +BoolInt CPU_IsSupported_PageGB() +{ + Cx86cpuid cpuid; + if (!x86cpuid_CheckAndRead(&cpuid)) + return False; + { + UInt32 d[4] = { 0 }; + MyCPUID(0x80000000, &d[0], &d[1], &d[2], &d[3]); + if (d[0] < 0x80000001) + return False; + } + { + UInt32 d[4] = { 0 }; + MyCPUID(0x80000001, &d[0], &d[1], &d[2], &d[3]); + return (d[3] >> 26) & 1; + } +} + + +#elif defined(MY_CPU_ARM_OR_ARM64) + +#ifdef _WIN32 + +#include + +BoolInt CPU_IsSupported_CRC32() { return IsProcessorFeaturePresent(PF_ARM_V8_CRC32_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } +BoolInt CPU_IsSupported_CRYPTO() { return IsProcessorFeaturePresent(PF_ARM_V8_CRYPTO_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } +BoolInt CPU_IsSupported_NEON() { return IsProcessorFeaturePresent(PF_ARM_NEON_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } + +#else + +#if defined(__APPLE__) + +/* +#include +#include +static void Print_sysctlbyname(const char *name) +{ + size_t bufSize = 256; + char buf[256]; + int res = sysctlbyname(name, &buf, &bufSize, NULL, 0); + { + int i; + printf("\nres = %d : %s : '%s' : bufSize = %d, numeric", res, name, buf, (unsigned)bufSize); + for (i = 0; i < 20; i++) + printf(" %2x", (unsigned)(Byte)buf[i]); + + } +} +*/ + +static BoolInt My_sysctlbyname_Get_BoolInt(const char *name) +{ + UInt32 val = 0; + if (My_sysctlbyname_Get_UInt32(name, &val) == 0 && val == 1) + return 1; + return 0; +} + + /* + Print_sysctlbyname("hw.pagesize"); + Print_sysctlbyname("machdep.cpu.brand_string"); + */ + +BoolInt CPU_IsSupported_CRC32(void) +{ + return My_sysctlbyname_Get_BoolInt("hw.optional.armv8_crc32"); +} + +BoolInt CPU_IsSupported_NEON(void) +{ + return My_sysctlbyname_Get_BoolInt("hw.optional.neon"); +} + +#ifdef MY_CPU_ARM64 +#define APPLE_CRYPTO_SUPPORT_VAL 1 +#else +#define APPLE_CRYPTO_SUPPORT_VAL 0 +#endif + +BoolInt CPU_IsSupported_SHA1(void) { return APPLE_CRYPTO_SUPPORT_VAL; } +BoolInt CPU_IsSupported_SHA2(void) { return APPLE_CRYPTO_SUPPORT_VAL; } +BoolInt CPU_IsSupported_AES (void) { return APPLE_CRYPTO_SUPPORT_VAL; } + + +#else // __APPLE__ + +#include + +#define USE_HWCAP + +#ifdef USE_HWCAP + +#include + + #define MY_HWCAP_CHECK_FUNC_2(name1, name2) \ + BoolInt CPU_IsSupported_ ## name1() { return (getauxval(AT_HWCAP) & (HWCAP_ ## name2)) ? 1 : 0; } + +#ifdef MY_CPU_ARM64 + #define MY_HWCAP_CHECK_FUNC(name) \ + MY_HWCAP_CHECK_FUNC_2(name, name) + MY_HWCAP_CHECK_FUNC_2(NEON, ASIMD) +// MY_HWCAP_CHECK_FUNC (ASIMD) +#elif defined(MY_CPU_ARM) + #define MY_HWCAP_CHECK_FUNC(name) \ + BoolInt CPU_IsSupported_ ## name() { return (getauxval(AT_HWCAP2) & (HWCAP2_ ## name)) ? 1 : 0; } + MY_HWCAP_CHECK_FUNC_2(NEON, NEON) +#endif + +#else // USE_HWCAP + + #define MY_HWCAP_CHECK_FUNC(name) \ + BoolInt CPU_IsSupported_ ## name() { return 0; } + MY_HWCAP_CHECK_FUNC(NEON) + +#endif // USE_HWCAP + +MY_HWCAP_CHECK_FUNC (CRC32) +MY_HWCAP_CHECK_FUNC (SHA1) +MY_HWCAP_CHECK_FUNC (SHA2) +MY_HWCAP_CHECK_FUNC (AES) + +#endif // __APPLE__ +#endif // _WIN32 + +#endif // MY_CPU_ARM_OR_ARM64 + + + +#ifdef __APPLE__ + +#include + +int My_sysctlbyname_Get(const char *name, void *buf, size_t *bufSize) +{ + return sysctlbyname(name, buf, bufSize, NULL, 0); +} + +int My_sysctlbyname_Get_UInt32(const char *name, UInt32 *val) +{ + size_t bufSize = sizeof(*val); + int res = My_sysctlbyname_Get(name, val, &bufSize); + if (res == 0 && bufSize != sizeof(*val)) + return EFAULT; + return res; +} + +#endif diff --git a/deps/LZMA-SDK/C/CpuArch.h b/deps/LZMA-SDK/C/CpuArch.h index ba8782714..4856fbb12 100644 --- a/deps/LZMA-SDK/C/CpuArch.h +++ b/deps/LZMA-SDK/C/CpuArch.h @@ -1,445 +1,445 @@ -/* CpuArch.h -- CPU specific code -2022-07-15 : Igor Pavlov : Public domain */ - -#ifndef __CPU_ARCH_H -#define __CPU_ARCH_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -/* -MY_CPU_LE means that CPU is LITTLE ENDIAN. -MY_CPU_BE means that CPU is BIG ENDIAN. -If MY_CPU_LE and MY_CPU_BE are not defined, we don't know about ENDIANNESS of platform. - -MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned memory accesses. - -MY_CPU_64BIT means that processor can work with 64-bit registers. - MY_CPU_64BIT can be used to select fast code branch - MY_CPU_64BIT doesn't mean that (sizeof(void *) == 8) -*/ - -#if defined(_M_X64) \ - || defined(_M_AMD64) \ - || defined(__x86_64__) \ - || defined(__AMD64__) \ - || defined(__amd64__) - #define MY_CPU_AMD64 - #ifdef __ILP32__ - #define MY_CPU_NAME "x32" - #define MY_CPU_SIZEOF_POINTER 4 - #else - #define MY_CPU_NAME "x64" - #define MY_CPU_SIZEOF_POINTER 8 - #endif - #define MY_CPU_64BIT -#endif - - -#if defined(_M_IX86) \ - || defined(__i386__) - #define MY_CPU_X86 - #define MY_CPU_NAME "x86" - /* #define MY_CPU_32BIT */ - #define MY_CPU_SIZEOF_POINTER 4 -#endif - - -#if defined(_M_ARM64) \ - || defined(__AARCH64EL__) \ - || defined(__AARCH64EB__) \ - || defined(__aarch64__) - #define MY_CPU_ARM64 - #define MY_CPU_NAME "arm64" - #define MY_CPU_64BIT -#endif - - -#if defined(_M_ARM) \ - || defined(_M_ARM_NT) \ - || defined(_M_ARMT) \ - || defined(__arm__) \ - || defined(__thumb__) \ - || defined(__ARMEL__) \ - || defined(__ARMEB__) \ - || defined(__THUMBEL__) \ - || defined(__THUMBEB__) - #define MY_CPU_ARM - - #if defined(__thumb__) || defined(__THUMBEL__) || defined(_M_ARMT) - #define MY_CPU_NAME "armt" - #else - #define MY_CPU_NAME "arm" - #endif - /* #define MY_CPU_32BIT */ - #define MY_CPU_SIZEOF_POINTER 4 -#endif - - -#if defined(_M_IA64) \ - || defined(__ia64__) - #define MY_CPU_IA64 - #define MY_CPU_NAME "ia64" - #define MY_CPU_64BIT -#endif - - -#if defined(__mips64) \ - || defined(__mips64__) \ - || (defined(__mips) && (__mips == 64 || __mips == 4 || __mips == 3)) - #define MY_CPU_NAME "mips64" - #define MY_CPU_64BIT -#elif defined(__mips__) - #define MY_CPU_NAME "mips" - /* #define MY_CPU_32BIT */ -#endif - - -#if defined(__ppc64__) \ - || defined(__powerpc64__) \ - || defined(__ppc__) \ - || defined(__powerpc__) \ - || defined(__PPC__) \ - || defined(_POWER) - -#if defined(__ppc64__) \ - || defined(__powerpc64__) \ - || defined(_LP64) \ - || defined(__64BIT__) - #ifdef __ILP32__ - #define MY_CPU_NAME "ppc64-32" - #define MY_CPU_SIZEOF_POINTER 4 - #else - #define MY_CPU_NAME "ppc64" - #define MY_CPU_SIZEOF_POINTER 8 - #endif - #define MY_CPU_64BIT -#else - #define MY_CPU_NAME "ppc" - #define MY_CPU_SIZEOF_POINTER 4 - /* #define MY_CPU_32BIT */ -#endif -#endif - - -#if defined(__riscv) \ - || defined(__riscv__) - #if __riscv_xlen == 32 - #define MY_CPU_NAME "riscv32" - #elif __riscv_xlen == 64 - #define MY_CPU_NAME "riscv64" - #else - #define MY_CPU_NAME "riscv" - #endif -#endif - - -#if defined(MY_CPU_X86) || defined(MY_CPU_AMD64) -#define MY_CPU_X86_OR_AMD64 -#endif - -#if defined(MY_CPU_ARM) || defined(MY_CPU_ARM64) -#define MY_CPU_ARM_OR_ARM64 -#endif - - -#ifdef _WIN32 - - #ifdef MY_CPU_ARM - #define MY_CPU_ARM_LE - #endif - - #ifdef MY_CPU_ARM64 - #define MY_CPU_ARM64_LE - #endif - - #ifdef _M_IA64 - #define MY_CPU_IA64_LE - #endif - -#endif - - -#if defined(MY_CPU_X86_OR_AMD64) \ - || defined(MY_CPU_ARM_LE) \ - || defined(MY_CPU_ARM64_LE) \ - || defined(MY_CPU_IA64_LE) \ - || defined(__LITTLE_ENDIAN__) \ - || defined(__ARMEL__) \ - || defined(__THUMBEL__) \ - || defined(__AARCH64EL__) \ - || defined(__MIPSEL__) \ - || defined(__MIPSEL) \ - || defined(_MIPSEL) \ - || defined(__BFIN__) \ - || (defined(__BYTE_ORDER__) && (__BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__)) - #define MY_CPU_LE -#endif - -#if defined(__BIG_ENDIAN__) \ - || defined(__ARMEB__) \ - || defined(__THUMBEB__) \ - || defined(__AARCH64EB__) \ - || defined(__MIPSEB__) \ - || defined(__MIPSEB) \ - || defined(_MIPSEB) \ - || defined(__m68k__) \ - || defined(__s390__) \ - || defined(__s390x__) \ - || defined(__zarch__) \ - || (defined(__BYTE_ORDER__) && (__BYTE_ORDER__ == __ORDER_BIG_ENDIAN__)) - #define MY_CPU_BE -#endif - - -#if defined(MY_CPU_LE) && defined(MY_CPU_BE) - #error Stop_Compiling_Bad_Endian -#endif - - -#if defined(MY_CPU_32BIT) && defined(MY_CPU_64BIT) - #error Stop_Compiling_Bad_32_64_BIT -#endif - -#ifdef __SIZEOF_POINTER__ - #ifdef MY_CPU_SIZEOF_POINTER - #if MY_CPU_SIZEOF_POINTER != __SIZEOF_POINTER__ - #error Stop_Compiling_Bad_MY_CPU_PTR_SIZE - #endif - #else - #define MY_CPU_SIZEOF_POINTER __SIZEOF_POINTER__ - #endif -#endif - -#if defined(MY_CPU_SIZEOF_POINTER) && (MY_CPU_SIZEOF_POINTER == 4) -#if defined (_LP64) - #error Stop_Compiling_Bad_MY_CPU_PTR_SIZE -#endif -#endif - -#ifdef _MSC_VER - #if _MSC_VER >= 1300 - #define MY_CPU_pragma_pack_push_1 __pragma(pack(push, 1)) - #define MY_CPU_pragma_pop __pragma(pack(pop)) - #else - #define MY_CPU_pragma_pack_push_1 - #define MY_CPU_pragma_pop - #endif -#else - #ifdef __xlC__ - #define MY_CPU_pragma_pack_push_1 _Pragma("pack(1)") - #define MY_CPU_pragma_pop _Pragma("pack()") - #else - #define MY_CPU_pragma_pack_push_1 _Pragma("pack(push, 1)") - #define MY_CPU_pragma_pop _Pragma("pack(pop)") - #endif -#endif - - -#ifndef MY_CPU_NAME - #ifdef MY_CPU_LE - #define MY_CPU_NAME "LE" - #elif defined(MY_CPU_BE) - #define MY_CPU_NAME "BE" - #else - /* - #define MY_CPU_NAME "" - */ - #endif -#endif - - - - - -#ifdef MY_CPU_LE - #if defined(MY_CPU_X86_OR_AMD64) \ - || defined(MY_CPU_ARM64) - #define MY_CPU_LE_UNALIGN - #define MY_CPU_LE_UNALIGN_64 - #elif defined(__ARM_FEATURE_UNALIGNED) - /* gcc9 for 32-bit arm can use LDRD instruction that requires 32-bit alignment. - So we can't use unaligned 64-bit operations. */ - #define MY_CPU_LE_UNALIGN - #endif -#endif - - -#ifdef MY_CPU_LE_UNALIGN - -#define GetUi16(p) (*(const UInt16 *)(const void *)(p)) -#define GetUi32(p) (*(const UInt32 *)(const void *)(p)) -#ifdef MY_CPU_LE_UNALIGN_64 -#define GetUi64(p) (*(const UInt64 *)(const void *)(p)) -#endif - -#define SetUi16(p, v) { *(UInt16 *)(void *)(p) = (v); } -#define SetUi32(p, v) { *(UInt32 *)(void *)(p) = (v); } -#ifdef MY_CPU_LE_UNALIGN_64 -#define SetUi64(p, v) { *(UInt64 *)(void *)(p) = (v); } -#endif - -#else - -#define GetUi16(p) ( (UInt16) ( \ - ((const Byte *)(p))[0] | \ - ((UInt16)((const Byte *)(p))[1] << 8) )) - -#define GetUi32(p) ( \ - ((const Byte *)(p))[0] | \ - ((UInt32)((const Byte *)(p))[1] << 8) | \ - ((UInt32)((const Byte *)(p))[2] << 16) | \ - ((UInt32)((const Byte *)(p))[3] << 24)) - -#define SetUi16(p, v) { Byte *_ppp_ = (Byte *)(p); UInt32 _vvv_ = (v); \ - _ppp_[0] = (Byte)_vvv_; \ - _ppp_[1] = (Byte)(_vvv_ >> 8); } - -#define SetUi32(p, v) { Byte *_ppp_ = (Byte *)(p); UInt32 _vvv_ = (v); \ - _ppp_[0] = (Byte)_vvv_; \ - _ppp_[1] = (Byte)(_vvv_ >> 8); \ - _ppp_[2] = (Byte)(_vvv_ >> 16); \ - _ppp_[3] = (Byte)(_vvv_ >> 24); } - -#endif - - -#ifndef MY_CPU_LE_UNALIGN_64 - -#define GetUi64(p) (GetUi32(p) | ((UInt64)GetUi32(((const Byte *)(p)) + 4) << 32)) - -#define SetUi64(p, v) { Byte *_ppp2_ = (Byte *)(p); UInt64 _vvv2_ = (v); \ - SetUi32(_ppp2_ , (UInt32)_vvv2_); \ - SetUi32(_ppp2_ + 4, (UInt32)(_vvv2_ >> 32)); } - -#endif - - - - -#ifdef __has_builtin - #define MY__has_builtin(x) __has_builtin(x) -#else - #define MY__has_builtin(x) 0 -#endif - -#if defined(MY_CPU_LE_UNALIGN) && /* defined(_WIN64) && */ defined(_MSC_VER) && (_MSC_VER >= 1300) - -/* Note: we use bswap instruction, that is unsupported in 386 cpu */ - -#include - -#pragma intrinsic(_byteswap_ushort) -#pragma intrinsic(_byteswap_ulong) -#pragma intrinsic(_byteswap_uint64) - -/* #define GetBe16(p) _byteswap_ushort(*(const UInt16 *)(const Byte *)(p)) */ -#define GetBe32(p) _byteswap_ulong (*(const UInt32 *)(const void *)(p)) -#define GetBe64(p) _byteswap_uint64(*(const UInt64 *)(const void *)(p)) - -#define SetBe32(p, v) (*(UInt32 *)(void *)(p)) = _byteswap_ulong(v) - -#elif defined(MY_CPU_LE_UNALIGN) && ( \ - (defined(__GNUC__) && (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 3))) \ - || (defined(__clang__) && MY__has_builtin(__builtin_bswap16)) ) - -/* #define GetBe16(p) __builtin_bswap16(*(const UInt16 *)(const void *)(p)) */ -#define GetBe32(p) __builtin_bswap32(*(const UInt32 *)(const void *)(p)) -#define GetBe64(p) __builtin_bswap64(*(const UInt64 *)(const void *)(p)) - -#define SetBe32(p, v) (*(UInt32 *)(void *)(p)) = __builtin_bswap32(v) - -#else - -#define GetBe32(p) ( \ - ((UInt32)((const Byte *)(p))[0] << 24) | \ - ((UInt32)((const Byte *)(p))[1] << 16) | \ - ((UInt32)((const Byte *)(p))[2] << 8) | \ - ((const Byte *)(p))[3] ) - -#define GetBe64(p) (((UInt64)GetBe32(p) << 32) | GetBe32(((const Byte *)(p)) + 4)) - -#define SetBe32(p, v) { Byte *_ppp_ = (Byte *)(p); UInt32 _vvv_ = (v); \ - _ppp_[0] = (Byte)(_vvv_ >> 24); \ - _ppp_[1] = (Byte)(_vvv_ >> 16); \ - _ppp_[2] = (Byte)(_vvv_ >> 8); \ - _ppp_[3] = (Byte)_vvv_; } - -#endif - - -#ifndef GetBe16 - -#define GetBe16(p) ( (UInt16) ( \ - ((UInt16)((const Byte *)(p))[0] << 8) | \ - ((const Byte *)(p))[1] )) - -#endif - - - -#ifdef MY_CPU_X86_OR_AMD64 - -typedef struct -{ - UInt32 maxFunc; - UInt32 vendor[3]; - UInt32 ver; - UInt32 b; - UInt32 c; - UInt32 d; -} Cx86cpuid; - -enum -{ - CPU_FIRM_INTEL, - CPU_FIRM_AMD, - CPU_FIRM_VIA -}; - -void MyCPUID(UInt32 function, UInt32 *a, UInt32 *b, UInt32 *c, UInt32 *d); - -BoolInt x86cpuid_CheckAndRead(Cx86cpuid *p); -int x86cpuid_GetFirm(const Cx86cpuid *p); - -#define x86cpuid_GetFamily(ver) (((ver >> 16) & 0xFF0) | ((ver >> 8) & 0xF)) -#define x86cpuid_GetModel(ver) (((ver >> 12) & 0xF0) | ((ver >> 4) & 0xF)) -#define x86cpuid_GetStepping(ver) (ver & 0xF) - -BoolInt CPU_Is_InOrder(void); - -BoolInt CPU_IsSupported_AES(void); -BoolInt CPU_IsSupported_AVX2(void); -BoolInt CPU_IsSupported_VAES_AVX2(void); -BoolInt CPU_IsSupported_SSSE3(void); -BoolInt CPU_IsSupported_SSE41(void); -BoolInt CPU_IsSupported_SHA(void); -BoolInt CPU_IsSupported_PageGB(void); - -#elif defined(MY_CPU_ARM_OR_ARM64) - -BoolInt CPU_IsSupported_CRC32(void); -BoolInt CPU_IsSupported_NEON(void); - -#if defined(_WIN32) -BoolInt CPU_IsSupported_CRYPTO(void); -#define CPU_IsSupported_SHA1 CPU_IsSupported_CRYPTO -#define CPU_IsSupported_SHA2 CPU_IsSupported_CRYPTO -#define CPU_IsSupported_AES CPU_IsSupported_CRYPTO -#else -BoolInt CPU_IsSupported_SHA1(void); -BoolInt CPU_IsSupported_SHA2(void); -BoolInt CPU_IsSupported_AES(void); -#endif - -#endif - -#if defined(__APPLE__) -int My_sysctlbyname_Get(const char *name, void *buf, size_t *bufSize); -int My_sysctlbyname_Get_UInt32(const char *name, UInt32 *val); -#endif - -EXTERN_C_END - -#endif +/* CpuArch.h -- CPU specific code +2022-07-15 : Igor Pavlov : Public domain */ + +#ifndef __CPU_ARCH_H +#define __CPU_ARCH_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +/* +MY_CPU_LE means that CPU is LITTLE ENDIAN. +MY_CPU_BE means that CPU is BIG ENDIAN. +If MY_CPU_LE and MY_CPU_BE are not defined, we don't know about ENDIANNESS of platform. + +MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned memory accesses. + +MY_CPU_64BIT means that processor can work with 64-bit registers. + MY_CPU_64BIT can be used to select fast code branch + MY_CPU_64BIT doesn't mean that (sizeof(void *) == 8) +*/ + +#if defined(_M_X64) \ + || defined(_M_AMD64) \ + || defined(__x86_64__) \ + || defined(__AMD64__) \ + || defined(__amd64__) + #define MY_CPU_AMD64 + #ifdef __ILP32__ + #define MY_CPU_NAME "x32" + #define MY_CPU_SIZEOF_POINTER 4 + #else + #define MY_CPU_NAME "x64" + #define MY_CPU_SIZEOF_POINTER 8 + #endif + #define MY_CPU_64BIT +#endif + + +#if defined(_M_IX86) \ + || defined(__i386__) + #define MY_CPU_X86 + #define MY_CPU_NAME "x86" + /* #define MY_CPU_32BIT */ + #define MY_CPU_SIZEOF_POINTER 4 +#endif + + +#if defined(_M_ARM64) \ + || defined(__AARCH64EL__) \ + || defined(__AARCH64EB__) \ + || defined(__aarch64__) + #define MY_CPU_ARM64 + #define MY_CPU_NAME "arm64" + #define MY_CPU_64BIT +#endif + + +#if defined(_M_ARM) \ + || defined(_M_ARM_NT) \ + || defined(_M_ARMT) \ + || defined(__arm__) \ + || defined(__thumb__) \ + || defined(__ARMEL__) \ + || defined(__ARMEB__) \ + || defined(__THUMBEL__) \ + || defined(__THUMBEB__) + #define MY_CPU_ARM + + #if defined(__thumb__) || defined(__THUMBEL__) || defined(_M_ARMT) + #define MY_CPU_NAME "armt" + #else + #define MY_CPU_NAME "arm" + #endif + /* #define MY_CPU_32BIT */ + #define MY_CPU_SIZEOF_POINTER 4 +#endif + + +#if defined(_M_IA64) \ + || defined(__ia64__) + #define MY_CPU_IA64 + #define MY_CPU_NAME "ia64" + #define MY_CPU_64BIT +#endif + + +#if defined(__mips64) \ + || defined(__mips64__) \ + || (defined(__mips) && (__mips == 64 || __mips == 4 || __mips == 3)) + #define MY_CPU_NAME "mips64" + #define MY_CPU_64BIT +#elif defined(__mips__) + #define MY_CPU_NAME "mips" + /* #define MY_CPU_32BIT */ +#endif + + +#if defined(__ppc64__) \ + || defined(__powerpc64__) \ + || defined(__ppc__) \ + || defined(__powerpc__) \ + || defined(__PPC__) \ + || defined(_POWER) + +#if defined(__ppc64__) \ + || defined(__powerpc64__) \ + || defined(_LP64) \ + || defined(__64BIT__) + #ifdef __ILP32__ + #define MY_CPU_NAME "ppc64-32" + #define MY_CPU_SIZEOF_POINTER 4 + #else + #define MY_CPU_NAME "ppc64" + #define MY_CPU_SIZEOF_POINTER 8 + #endif + #define MY_CPU_64BIT +#else + #define MY_CPU_NAME "ppc" + #define MY_CPU_SIZEOF_POINTER 4 + /* #define MY_CPU_32BIT */ +#endif +#endif + + +#if defined(__riscv) \ + || defined(__riscv__) + #if __riscv_xlen == 32 + #define MY_CPU_NAME "riscv32" + #elif __riscv_xlen == 64 + #define MY_CPU_NAME "riscv64" + #else + #define MY_CPU_NAME "riscv" + #endif +#endif + + +#if defined(MY_CPU_X86) || defined(MY_CPU_AMD64) +#define MY_CPU_X86_OR_AMD64 +#endif + +#if defined(MY_CPU_ARM) || defined(MY_CPU_ARM64) +#define MY_CPU_ARM_OR_ARM64 +#endif + + +#ifdef _WIN32 + + #ifdef MY_CPU_ARM + #define MY_CPU_ARM_LE + #endif + + #ifdef MY_CPU_ARM64 + #define MY_CPU_ARM64_LE + #endif + + #ifdef _M_IA64 + #define MY_CPU_IA64_LE + #endif + +#endif + + +#if defined(MY_CPU_X86_OR_AMD64) \ + || defined(MY_CPU_ARM_LE) \ + || defined(MY_CPU_ARM64_LE) \ + || defined(MY_CPU_IA64_LE) \ + || defined(__LITTLE_ENDIAN__) \ + || defined(__ARMEL__) \ + || defined(__THUMBEL__) \ + || defined(__AARCH64EL__) \ + || defined(__MIPSEL__) \ + || defined(__MIPSEL) \ + || defined(_MIPSEL) \ + || defined(__BFIN__) \ + || (defined(__BYTE_ORDER__) && (__BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__)) + #define MY_CPU_LE +#endif + +#if defined(__BIG_ENDIAN__) \ + || defined(__ARMEB__) \ + || defined(__THUMBEB__) \ + || defined(__AARCH64EB__) \ + || defined(__MIPSEB__) \ + || defined(__MIPSEB) \ + || defined(_MIPSEB) \ + || defined(__m68k__) \ + || defined(__s390__) \ + || defined(__s390x__) \ + || defined(__zarch__) \ + || (defined(__BYTE_ORDER__) && (__BYTE_ORDER__ == __ORDER_BIG_ENDIAN__)) + #define MY_CPU_BE +#endif + + +#if defined(MY_CPU_LE) && defined(MY_CPU_BE) + #error Stop_Compiling_Bad_Endian +#endif + + +#if defined(MY_CPU_32BIT) && defined(MY_CPU_64BIT) + #error Stop_Compiling_Bad_32_64_BIT +#endif + +#ifdef __SIZEOF_POINTER__ + #ifdef MY_CPU_SIZEOF_POINTER + #if MY_CPU_SIZEOF_POINTER != __SIZEOF_POINTER__ + #error Stop_Compiling_Bad_MY_CPU_PTR_SIZE + #endif + #else + #define MY_CPU_SIZEOF_POINTER __SIZEOF_POINTER__ + #endif +#endif + +#if defined(MY_CPU_SIZEOF_POINTER) && (MY_CPU_SIZEOF_POINTER == 4) +#if defined (_LP64) + #error Stop_Compiling_Bad_MY_CPU_PTR_SIZE +#endif +#endif + +#ifdef _MSC_VER + #if _MSC_VER >= 1300 + #define MY_CPU_pragma_pack_push_1 __pragma(pack(push, 1)) + #define MY_CPU_pragma_pop __pragma(pack(pop)) + #else + #define MY_CPU_pragma_pack_push_1 + #define MY_CPU_pragma_pop + #endif +#else + #ifdef __xlC__ + #define MY_CPU_pragma_pack_push_1 _Pragma("pack(1)") + #define MY_CPU_pragma_pop _Pragma("pack()") + #else + #define MY_CPU_pragma_pack_push_1 _Pragma("pack(push, 1)") + #define MY_CPU_pragma_pop _Pragma("pack(pop)") + #endif +#endif + + +#ifndef MY_CPU_NAME + #ifdef MY_CPU_LE + #define MY_CPU_NAME "LE" + #elif defined(MY_CPU_BE) + #define MY_CPU_NAME "BE" + #else + /* + #define MY_CPU_NAME "" + */ + #endif +#endif + + + + + +#ifdef MY_CPU_LE + #if defined(MY_CPU_X86_OR_AMD64) \ + || defined(MY_CPU_ARM64) + #define MY_CPU_LE_UNALIGN + #define MY_CPU_LE_UNALIGN_64 + #elif defined(__ARM_FEATURE_UNALIGNED) + /* gcc9 for 32-bit arm can use LDRD instruction that requires 32-bit alignment. + So we can't use unaligned 64-bit operations. */ + #define MY_CPU_LE_UNALIGN + #endif +#endif + + +#ifdef MY_CPU_LE_UNALIGN + +#define GetUi16(p) (*(const UInt16 *)(const void *)(p)) +#define GetUi32(p) (*(const UInt32 *)(const void *)(p)) +#ifdef MY_CPU_LE_UNALIGN_64 +#define GetUi64(p) (*(const UInt64 *)(const void *)(p)) +#endif + +#define SetUi16(p, v) { *(UInt16 *)(void *)(p) = (v); } +#define SetUi32(p, v) { *(UInt32 *)(void *)(p) = (v); } +#ifdef MY_CPU_LE_UNALIGN_64 +#define SetUi64(p, v) { *(UInt64 *)(void *)(p) = (v); } +#endif + +#else + +#define GetUi16(p) ( (UInt16) ( \ + ((const Byte *)(p))[0] | \ + ((UInt16)((const Byte *)(p))[1] << 8) )) + +#define GetUi32(p) ( \ + ((const Byte *)(p))[0] | \ + ((UInt32)((const Byte *)(p))[1] << 8) | \ + ((UInt32)((const Byte *)(p))[2] << 16) | \ + ((UInt32)((const Byte *)(p))[3] << 24)) + +#define SetUi16(p, v) { Byte *_ppp_ = (Byte *)(p); UInt32 _vvv_ = (v); \ + _ppp_[0] = (Byte)_vvv_; \ + _ppp_[1] = (Byte)(_vvv_ >> 8); } + +#define SetUi32(p, v) { Byte *_ppp_ = (Byte *)(p); UInt32 _vvv_ = (v); \ + _ppp_[0] = (Byte)_vvv_; \ + _ppp_[1] = (Byte)(_vvv_ >> 8); \ + _ppp_[2] = (Byte)(_vvv_ >> 16); \ + _ppp_[3] = (Byte)(_vvv_ >> 24); } + +#endif + + +#ifndef MY_CPU_LE_UNALIGN_64 + +#define GetUi64(p) (GetUi32(p) | ((UInt64)GetUi32(((const Byte *)(p)) + 4) << 32)) + +#define SetUi64(p, v) { Byte *_ppp2_ = (Byte *)(p); UInt64 _vvv2_ = (v); \ + SetUi32(_ppp2_ , (UInt32)_vvv2_); \ + SetUi32(_ppp2_ + 4, (UInt32)(_vvv2_ >> 32)); } + +#endif + + + + +#ifdef __has_builtin + #define MY__has_builtin(x) __has_builtin(x) +#else + #define MY__has_builtin(x) 0 +#endif + +#if defined(MY_CPU_LE_UNALIGN) && /* defined(_WIN64) && */ defined(_MSC_VER) && (_MSC_VER >= 1300) + +/* Note: we use bswap instruction, that is unsupported in 386 cpu */ + +#include + +#pragma intrinsic(_byteswap_ushort) +#pragma intrinsic(_byteswap_ulong) +#pragma intrinsic(_byteswap_uint64) + +/* #define GetBe16(p) _byteswap_ushort(*(const UInt16 *)(const Byte *)(p)) */ +#define GetBe32(p) _byteswap_ulong (*(const UInt32 *)(const void *)(p)) +#define GetBe64(p) _byteswap_uint64(*(const UInt64 *)(const void *)(p)) + +#define SetBe32(p, v) (*(UInt32 *)(void *)(p)) = _byteswap_ulong(v) + +#elif defined(MY_CPU_LE_UNALIGN) && ( \ + (defined(__GNUC__) && (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 3))) \ + || (defined(__clang__) && MY__has_builtin(__builtin_bswap16)) ) + +/* #define GetBe16(p) __builtin_bswap16(*(const UInt16 *)(const void *)(p)) */ +#define GetBe32(p) __builtin_bswap32(*(const UInt32 *)(const void *)(p)) +#define GetBe64(p) __builtin_bswap64(*(const UInt64 *)(const void *)(p)) + +#define SetBe32(p, v) (*(UInt32 *)(void *)(p)) = __builtin_bswap32(v) + +#else + +#define GetBe32(p) ( \ + ((UInt32)((const Byte *)(p))[0] << 24) | \ + ((UInt32)((const Byte *)(p))[1] << 16) | \ + ((UInt32)((const Byte *)(p))[2] << 8) | \ + ((const Byte *)(p))[3] ) + +#define GetBe64(p) (((UInt64)GetBe32(p) << 32) | GetBe32(((const Byte *)(p)) + 4)) + +#define SetBe32(p, v) { Byte *_ppp_ = (Byte *)(p); UInt32 _vvv_ = (v); \ + _ppp_[0] = (Byte)(_vvv_ >> 24); \ + _ppp_[1] = (Byte)(_vvv_ >> 16); \ + _ppp_[2] = (Byte)(_vvv_ >> 8); \ + _ppp_[3] = (Byte)_vvv_; } + +#endif + + +#ifndef GetBe16 + +#define GetBe16(p) ( (UInt16) ( \ + ((UInt16)((const Byte *)(p))[0] << 8) | \ + ((const Byte *)(p))[1] )) + +#endif + + + +#ifdef MY_CPU_X86_OR_AMD64 + +typedef struct +{ + UInt32 maxFunc; + UInt32 vendor[3]; + UInt32 ver; + UInt32 b; + UInt32 c; + UInt32 d; +} Cx86cpuid; + +enum +{ + CPU_FIRM_INTEL, + CPU_FIRM_AMD, + CPU_FIRM_VIA +}; + +void MyCPUID(UInt32 function, UInt32 *a, UInt32 *b, UInt32 *c, UInt32 *d); + +BoolInt x86cpuid_CheckAndRead(Cx86cpuid *p); +int x86cpuid_GetFirm(const Cx86cpuid *p); + +#define x86cpuid_GetFamily(ver) (((ver >> 16) & 0xFF0) | ((ver >> 8) & 0xF)) +#define x86cpuid_GetModel(ver) (((ver >> 12) & 0xF0) | ((ver >> 4) & 0xF)) +#define x86cpuid_GetStepping(ver) (ver & 0xF) + +BoolInt CPU_Is_InOrder(void); + +BoolInt CPU_IsSupported_AES(void); +BoolInt CPU_IsSupported_AVX2(void); +BoolInt CPU_IsSupported_VAES_AVX2(void); +BoolInt CPU_IsSupported_SSSE3(void); +BoolInt CPU_IsSupported_SSE41(void); +BoolInt CPU_IsSupported_SHA(void); +BoolInt CPU_IsSupported_PageGB(void); + +#elif defined(MY_CPU_ARM_OR_ARM64) + +BoolInt CPU_IsSupported_CRC32(void); +BoolInt CPU_IsSupported_NEON(void); + +#if defined(_WIN32) +BoolInt CPU_IsSupported_CRYPTO(void); +#define CPU_IsSupported_SHA1 CPU_IsSupported_CRYPTO +#define CPU_IsSupported_SHA2 CPU_IsSupported_CRYPTO +#define CPU_IsSupported_AES CPU_IsSupported_CRYPTO +#else +BoolInt CPU_IsSupported_SHA1(void); +BoolInt CPU_IsSupported_SHA2(void); +BoolInt CPU_IsSupported_AES(void); +#endif + +#endif + +#if defined(__APPLE__) +int My_sysctlbyname_Get(const char *name, void *buf, size_t *bufSize); +int My_sysctlbyname_Get_UInt32(const char *name, UInt32 *val); +#endif + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/Delta.c b/deps/LZMA-SDK/C/Delta.c index fc7e9fe96..c4a4499fe 100644 --- a/deps/LZMA-SDK/C/Delta.c +++ b/deps/LZMA-SDK/C/Delta.c @@ -1,169 +1,169 @@ -/* Delta.c -- Delta converter -2021-02-09 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "Delta.h" - -void Delta_Init(Byte *state) -{ - unsigned i; - for (i = 0; i < DELTA_STATE_SIZE; i++) - state[i] = 0; -} - - -void Delta_Encode(Byte *state, unsigned delta, Byte *data, SizeT size) -{ - Byte temp[DELTA_STATE_SIZE]; - - if (size == 0) - return; - - { - unsigned i = 0; - do - temp[i] = state[i]; - while (++i != delta); - } - - if (size <= delta) - { - unsigned i = 0, k; - do - { - Byte b = *data; - *data++ = (Byte)(b - temp[i]); - temp[i] = b; - } - while (++i != size); - - k = 0; - - do - { - if (i == delta) - i = 0; - state[k] = temp[i++]; - } - while (++k != delta); - - return; - } - - { - Byte *p = data + size - delta; - { - unsigned i = 0; - do - state[i] = *p++; - while (++i != delta); - } - { - const Byte *lim = data + delta; - ptrdiff_t dif = -(ptrdiff_t)delta; - - if (((ptrdiff_t)size + dif) & 1) - { - --p; *p = (Byte)(*p - p[dif]); - } - - while (p != lim) - { - --p; *p = (Byte)(*p - p[dif]); - --p; *p = (Byte)(*p - p[dif]); - } - - dif = -dif; - - do - { - --p; *p = (Byte)(*p - temp[--dif]); - } - while (dif != 0); - } - } -} - - -void Delta_Decode(Byte *state, unsigned delta, Byte *data, SizeT size) -{ - unsigned i; - const Byte *lim; - - if (size == 0) - return; - - i = 0; - lim = data + size; - - if (size <= delta) - { - do - *data = (Byte)(*data + state[i++]); - while (++data != lim); - - for (; delta != i; state++, delta--) - *state = state[i]; - data -= i; - } - else - { - /* - #define B(n) b ## n - #define I(n) Byte B(n) = state[n]; - #define U(n) { B(n) = (Byte)((B(n)) + *data++); data[-1] = (B(n)); } - #define F(n) if (data != lim) { U(n) } - - if (delta == 1) - { - I(0) - if ((lim - data) & 1) { U(0) } - while (data != lim) { U(0) U(0) } - data -= 1; - } - else if (delta == 2) - { - I(0) I(1) - lim -= 1; while (data < lim) { U(0) U(1) } - lim += 1; F(0) - data -= 2; - } - else if (delta == 3) - { - I(0) I(1) I(2) - lim -= 2; while (data < lim) { U(0) U(1) U(2) } - lim += 2; F(0) F(1) - data -= 3; - } - else if (delta == 4) - { - I(0) I(1) I(2) I(3) - lim -= 3; while (data < lim) { U(0) U(1) U(2) U(3) } - lim += 3; F(0) F(1) F(2) - data -= 4; - } - else - */ - { - do - { - *data = (Byte)(*data + state[i++]); - data++; - } - while (i != delta); - - { - ptrdiff_t dif = -(ptrdiff_t)delta; - do - *data = (Byte)(*data + data[dif]); - while (++data != lim); - data += dif; - } - } - } - - do - *state++ = *data; - while (++data != lim); -} +/* Delta.c -- Delta converter +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "Delta.h" + +void Delta_Init(Byte *state) +{ + unsigned i; + for (i = 0; i < DELTA_STATE_SIZE; i++) + state[i] = 0; +} + + +void Delta_Encode(Byte *state, unsigned delta, Byte *data, SizeT size) +{ + Byte temp[DELTA_STATE_SIZE]; + + if (size == 0) + return; + + { + unsigned i = 0; + do + temp[i] = state[i]; + while (++i != delta); + } + + if (size <= delta) + { + unsigned i = 0, k; + do + { + Byte b = *data; + *data++ = (Byte)(b - temp[i]); + temp[i] = b; + } + while (++i != size); + + k = 0; + + do + { + if (i == delta) + i = 0; + state[k] = temp[i++]; + } + while (++k != delta); + + return; + } + + { + Byte *p = data + size - delta; + { + unsigned i = 0; + do + state[i] = *p++; + while (++i != delta); + } + { + const Byte *lim = data + delta; + ptrdiff_t dif = -(ptrdiff_t)delta; + + if (((ptrdiff_t)size + dif) & 1) + { + --p; *p = (Byte)(*p - p[dif]); + } + + while (p != lim) + { + --p; *p = (Byte)(*p - p[dif]); + --p; *p = (Byte)(*p - p[dif]); + } + + dif = -dif; + + do + { + --p; *p = (Byte)(*p - temp[--dif]); + } + while (dif != 0); + } + } +} + + +void Delta_Decode(Byte *state, unsigned delta, Byte *data, SizeT size) +{ + unsigned i; + const Byte *lim; + + if (size == 0) + return; + + i = 0; + lim = data + size; + + if (size <= delta) + { + do + *data = (Byte)(*data + state[i++]); + while (++data != lim); + + for (; delta != i; state++, delta--) + *state = state[i]; + data -= i; + } + else + { + /* + #define B(n) b ## n + #define I(n) Byte B(n) = state[n]; + #define U(n) { B(n) = (Byte)((B(n)) + *data++); data[-1] = (B(n)); } + #define F(n) if (data != lim) { U(n) } + + if (delta == 1) + { + I(0) + if ((lim - data) & 1) { U(0) } + while (data != lim) { U(0) U(0) } + data -= 1; + } + else if (delta == 2) + { + I(0) I(1) + lim -= 1; while (data < lim) { U(0) U(1) } + lim += 1; F(0) + data -= 2; + } + else if (delta == 3) + { + I(0) I(1) I(2) + lim -= 2; while (data < lim) { U(0) U(1) U(2) } + lim += 2; F(0) F(1) + data -= 3; + } + else if (delta == 4) + { + I(0) I(1) I(2) I(3) + lim -= 3; while (data < lim) { U(0) U(1) U(2) U(3) } + lim += 3; F(0) F(1) F(2) + data -= 4; + } + else + */ + { + do + { + *data = (Byte)(*data + state[i++]); + data++; + } + while (i != delta); + + { + ptrdiff_t dif = -(ptrdiff_t)delta; + do + *data = (Byte)(*data + data[dif]); + while (++data != lim); + data += dif; + } + } + } + + do + *state++ = *data; + while (++data != lim); +} diff --git a/deps/LZMA-SDK/C/Delta.h b/deps/LZMA-SDK/C/Delta.h index e59d5a252..2fa54ad67 100644 --- a/deps/LZMA-SDK/C/Delta.h +++ b/deps/LZMA-SDK/C/Delta.h @@ -1,19 +1,19 @@ -/* Delta.h -- Delta converter -2013-01-18 : Igor Pavlov : Public domain */ - -#ifndef __DELTA_H -#define __DELTA_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -#define DELTA_STATE_SIZE 256 - -void Delta_Init(Byte *state); -void Delta_Encode(Byte *state, unsigned delta, Byte *data, SizeT size); -void Delta_Decode(Byte *state, unsigned delta, Byte *data, SizeT size); - -EXTERN_C_END - -#endif +/* Delta.h -- Delta converter +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __DELTA_H +#define __DELTA_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define DELTA_STATE_SIZE 256 + +void Delta_Init(Byte *state); +void Delta_Encode(Byte *state, unsigned delta, Byte *data, SizeT size); +void Delta_Decode(Byte *state, unsigned delta, Byte *data, SizeT size); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/DllSecur.c b/deps/LZMA-SDK/C/DllSecur.c index 16755bba9..a3c294bfe 100644 --- a/deps/LZMA-SDK/C/DllSecur.c +++ b/deps/LZMA-SDK/C/DllSecur.c @@ -1,114 +1,114 @@ -/* DllSecur.c -- DLL loading security -2022-07-15 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#ifdef _WIN32 - -#include - -#include "DllSecur.h" - -#ifndef UNDER_CE - -#if defined(__GNUC__) && (__GNUC__ >= 8) - #pragma GCC diagnostic ignored "-Wcast-function-type" -#endif - -typedef BOOL (WINAPI *Func_SetDefaultDllDirectories)(DWORD DirectoryFlags); - -#define MY_LOAD_LIBRARY_SEARCH_USER_DIRS 0x400 -#define MY_LOAD_LIBRARY_SEARCH_SYSTEM32 0x800 - -static const char * const g_Dlls = - #ifndef _CONSOLE - "UXTHEME\0" - #endif - "USERENV\0" - "SETUPAPI\0" - "APPHELP\0" - "PROPSYS\0" - "DWMAPI\0" - "CRYPTBASE\0" - "OLEACC\0" - "CLBCATQ\0" - "VERSION\0" - ; - -#endif - -// #define MY_CAST_FUNC (void(*)()) -#define MY_CAST_FUNC - -void My_SetDefaultDllDirectories() -{ - #ifndef UNDER_CE - - OSVERSIONINFO vi; - vi.dwOSVersionInfoSize = sizeof(vi); - if (!GetVersionEx(&vi) || vi.dwMajorVersion != 6 || vi.dwMinorVersion != 0) - { - Func_SetDefaultDllDirectories setDllDirs = (Func_SetDefaultDllDirectories) - MY_CAST_FUNC GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "SetDefaultDllDirectories"); - if (setDllDirs) - if (setDllDirs(MY_LOAD_LIBRARY_SEARCH_SYSTEM32 | MY_LOAD_LIBRARY_SEARCH_USER_DIRS)) - return; - } - - #endif -} - - -void LoadSecurityDlls() -{ - #ifndef UNDER_CE - - wchar_t buf[MAX_PATH + 100]; - - { - // at Vista (ver 6.0) : CoCreateInstance(CLSID_ShellLink, ...) doesn't work after SetDefaultDllDirectories() : Check it ??? - OSVERSIONINFO vi; - vi.dwOSVersionInfoSize = sizeof(vi); - if (!GetVersionEx(&vi) || vi.dwMajorVersion != 6 || vi.dwMinorVersion != 0) - { - Func_SetDefaultDllDirectories setDllDirs = (Func_SetDefaultDllDirectories) - MY_CAST_FUNC GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "SetDefaultDllDirectories"); - if (setDllDirs) - if (setDllDirs(MY_LOAD_LIBRARY_SEARCH_SYSTEM32 | MY_LOAD_LIBRARY_SEARCH_USER_DIRS)) - return; - } - } - - { - unsigned len = GetSystemDirectoryW(buf, MAX_PATH + 2); - if (len == 0 || len > MAX_PATH) - return; - } - { - const char *dll; - unsigned pos = (unsigned)lstrlenW(buf); - - if (buf[pos - 1] != '\\') - buf[pos++] = '\\'; - - for (dll = g_Dlls; dll[0] != 0;) - { - unsigned k = 0; - for (;;) - { - char c = *dll++; - buf[pos + k] = (Byte)c; - k++; - if (c == 0) - break; - } - - lstrcatW(buf, L".dll"); - LoadLibraryExW(buf, NULL, LOAD_WITH_ALTERED_SEARCH_PATH); - } - } - - #endif -} - -#endif +/* DllSecur.c -- DLL loading security +2022-07-15 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#ifdef _WIN32 + +#include + +#include "DllSecur.h" + +#ifndef UNDER_CE + +#if defined(__GNUC__) && (__GNUC__ >= 8) + #pragma GCC diagnostic ignored "-Wcast-function-type" +#endif + +typedef BOOL (WINAPI *Func_SetDefaultDllDirectories)(DWORD DirectoryFlags); + +#define MY_LOAD_LIBRARY_SEARCH_USER_DIRS 0x400 +#define MY_LOAD_LIBRARY_SEARCH_SYSTEM32 0x800 + +static const char * const g_Dlls = + #ifndef _CONSOLE + "UXTHEME\0" + #endif + "USERENV\0" + "SETUPAPI\0" + "APPHELP\0" + "PROPSYS\0" + "DWMAPI\0" + "CRYPTBASE\0" + "OLEACC\0" + "CLBCATQ\0" + "VERSION\0" + ; + +#endif + +// #define MY_CAST_FUNC (void(*)()) +#define MY_CAST_FUNC + +void My_SetDefaultDllDirectories() +{ + #ifndef UNDER_CE + + OSVERSIONINFO vi; + vi.dwOSVersionInfoSize = sizeof(vi); + if (!GetVersionEx(&vi) || vi.dwMajorVersion != 6 || vi.dwMinorVersion != 0) + { + Func_SetDefaultDllDirectories setDllDirs = (Func_SetDefaultDllDirectories) + MY_CAST_FUNC GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "SetDefaultDllDirectories"); + if (setDllDirs) + if (setDllDirs(MY_LOAD_LIBRARY_SEARCH_SYSTEM32 | MY_LOAD_LIBRARY_SEARCH_USER_DIRS)) + return; + } + + #endif +} + + +void LoadSecurityDlls() +{ + #ifndef UNDER_CE + + wchar_t buf[MAX_PATH + 100]; + + { + // at Vista (ver 6.0) : CoCreateInstance(CLSID_ShellLink, ...) doesn't work after SetDefaultDllDirectories() : Check it ??? + OSVERSIONINFO vi; + vi.dwOSVersionInfoSize = sizeof(vi); + if (!GetVersionEx(&vi) || vi.dwMajorVersion != 6 || vi.dwMinorVersion != 0) + { + Func_SetDefaultDllDirectories setDllDirs = (Func_SetDefaultDllDirectories) + MY_CAST_FUNC GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "SetDefaultDllDirectories"); + if (setDllDirs) + if (setDllDirs(MY_LOAD_LIBRARY_SEARCH_SYSTEM32 | MY_LOAD_LIBRARY_SEARCH_USER_DIRS)) + return; + } + } + + { + unsigned len = GetSystemDirectoryW(buf, MAX_PATH + 2); + if (len == 0 || len > MAX_PATH) + return; + } + { + const char *dll; + unsigned pos = (unsigned)lstrlenW(buf); + + if (buf[pos - 1] != '\\') + buf[pos++] = '\\'; + + for (dll = g_Dlls; dll[0] != 0;) + { + unsigned k = 0; + for (;;) + { + char c = *dll++; + buf[pos + k] = (Byte)c; + k++; + if (c == 0) + break; + } + + lstrcatW(buf, L".dll"); + LoadLibraryExW(buf, NULL, LOAD_WITH_ALTERED_SEARCH_PATH); + } + } + + #endif +} + +#endif diff --git a/deps/LZMA-SDK/C/DllSecur.h b/deps/LZMA-SDK/C/DllSecur.h index 0fd8070e5..64ff26cd9 100644 --- a/deps/LZMA-SDK/C/DllSecur.h +++ b/deps/LZMA-SDK/C/DllSecur.h @@ -1,20 +1,20 @@ -/* DllSecur.h -- DLL loading for security -2018-02-19 : Igor Pavlov : Public domain */ - -#ifndef __DLL_SECUR_H -#define __DLL_SECUR_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -#ifdef _WIN32 - -void My_SetDefaultDllDirectories(void); -void LoadSecurityDlls(void); - -#endif - -EXTERN_C_END - -#endif +/* DllSecur.h -- DLL loading for security +2018-02-19 : Igor Pavlov : Public domain */ + +#ifndef __DLL_SECUR_H +#define __DLL_SECUR_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#ifdef _WIN32 + +void My_SetDefaultDllDirectories(void); +void LoadSecurityDlls(void); + +#endif + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/LzFind.c b/deps/LZMA-SDK/C/LzFind.c index a17c06bc3..1b73c2848 100644 --- a/deps/LZMA-SDK/C/LzFind.c +++ b/deps/LZMA-SDK/C/LzFind.c @@ -1,1628 +1,1628 @@ -/* LzFind.c -- Match finder for LZ algorithms -2021-11-29 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include -// #include - -#include "CpuArch.h" -#include "LzFind.h" -#include "LzHash.h" - -#define kBlockMoveAlign (1 << 7) // alignment for memmove() -#define kBlockSizeAlign (1 << 16) // alignment for block allocation -#define kBlockSizeReserveMin (1 << 24) // it's 1/256 from 4 GB dictinary - -#define kEmptyHashValue 0 - -#define kMaxValForNormalize ((UInt32)0) -// #define kMaxValForNormalize ((UInt32)(1 << 20) + 0xFFF) // for debug - -// #define kNormalizeAlign (1 << 7) // alignment for speculated accesses - -#define GET_AVAIL_BYTES(p) \ - Inline_MatchFinder_GetNumAvailableBytes(p) - - -// #define kFix5HashSize (kHash2Size + kHash3Size + kHash4Size) -#define kFix5HashSize kFix4HashSize - -/* - HASH2_CALC: - if (hv) match, then cur[0] and cur[1] also match -*/ -#define HASH2_CALC hv = GetUi16(cur); - -// (crc[0 ... 255] & 0xFF) provides one-to-one correspondence to [0 ... 255] - -/* - HASH3_CALC: - if (cur[0]) and (h2) match, then cur[1] also match - if (cur[0]) and (hv) match, then cur[1] and cur[2] also match -*/ -#define HASH3_CALC { \ - UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ - h2 = temp & (kHash2Size - 1); \ - hv = (temp ^ ((UInt32)cur[2] << 8)) & p->hashMask; } - -#define HASH4_CALC { \ - UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ - h2 = temp & (kHash2Size - 1); \ - temp ^= ((UInt32)cur[2] << 8); \ - h3 = temp & (kHash3Size - 1); \ - hv = (temp ^ (p->crc[cur[3]] << kLzHash_CrcShift_1)) & p->hashMask; } - -#define HASH5_CALC { \ - UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ - h2 = temp & (kHash2Size - 1); \ - temp ^= ((UInt32)cur[2] << 8); \ - h3 = temp & (kHash3Size - 1); \ - temp ^= (p->crc[cur[3]] << kLzHash_CrcShift_1); \ - /* h4 = temp & p->hash4Mask; */ /* (kHash4Size - 1); */ \ - hv = (temp ^ (p->crc[cur[4]] << kLzHash_CrcShift_2)) & p->hashMask; } - -#define HASH_ZIP_CALC hv = ((cur[2] | ((UInt32)cur[0] << 8)) ^ p->crc[cur[1]]) & 0xFFFF; - - -static void LzInWindow_Free(CMatchFinder *p, ISzAllocPtr alloc) -{ - if (!p->directInput) - { - ISzAlloc_Free(alloc, p->bufferBase); - p->bufferBase = NULL; - } -} - - -static int LzInWindow_Create2(CMatchFinder *p, UInt32 blockSize, ISzAllocPtr alloc) -{ - if (blockSize == 0) - return 0; - if (!p->bufferBase || p->blockSize != blockSize) - { - // size_t blockSizeT; - LzInWindow_Free(p, alloc); - p->blockSize = blockSize; - // blockSizeT = blockSize; - - // printf("\nblockSize = 0x%x\n", blockSize); - /* - #if defined _WIN64 - // we can allocate 4GiB, but still use UInt32 for (p->blockSize) - // we use UInt32 type for (p->blockSize), because - // we don't want to wrap over 4 GiB, - // when we use (p->streamPos - p->pos) that is UInt32. - if (blockSize >= (UInt32)0 - (UInt32)kBlockSizeAlign) - { - blockSizeT = ((size_t)1 << 32); - printf("\nchanged to blockSizeT = 4GiB\n"); - } - #endif - */ - - p->bufferBase = (Byte *)ISzAlloc_Alloc(alloc, blockSize); - // printf("\nbufferBase = %p\n", p->bufferBase); - // return 0; // for debug - } - return (p->bufferBase != NULL); -} - -static const Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p) { return p->buffer; } - -static UInt32 MatchFinder_GetNumAvailableBytes(CMatchFinder *p) { return GET_AVAIL_BYTES(p); } - - -MY_NO_INLINE -static void MatchFinder_ReadBlock(CMatchFinder *p) -{ - if (p->streamEndWasReached || p->result != SZ_OK) - return; - - /* We use (p->streamPos - p->pos) value. - (p->streamPos < p->pos) is allowed. */ - - if (p->directInput) - { - UInt32 curSize = 0xFFFFFFFF - GET_AVAIL_BYTES(p); - if (curSize > p->directInputRem) - curSize = (UInt32)p->directInputRem; - p->directInputRem -= curSize; - p->streamPos += curSize; - if (p->directInputRem == 0) - p->streamEndWasReached = 1; - return; - } - - for (;;) - { - Byte *dest = p->buffer + GET_AVAIL_BYTES(p); - size_t size = (size_t)(p->bufferBase + p->blockSize - dest); - if (size == 0) - { - /* we call ReadBlock() after NeedMove() and MoveBlock(). - NeedMove() and MoveBlock() povide more than (keepSizeAfter) - to the end of (blockSize). - So we don't execute this branch in normal code flow. - We can go here, if we will call ReadBlock() before NeedMove(), MoveBlock(). - */ - // p->result = SZ_ERROR_FAIL; // we can show error here - return; - } - - // #define kRead 3 - // if (size > kRead) size = kRead; // for debug - - p->result = ISeqInStream_Read(p->stream, dest, &size); - if (p->result != SZ_OK) - return; - if (size == 0) - { - p->streamEndWasReached = 1; - return; - } - p->streamPos += (UInt32)size; - if (GET_AVAIL_BYTES(p) > p->keepSizeAfter) - return; - /* here and in another (p->keepSizeAfter) checks we keep on 1 byte more than was requested by Create() function - (GET_AVAIL_BYTES(p) >= p->keepSizeAfter) - minimal required size */ - } - - // on exit: (p->result != SZ_OK || p->streamEndWasReached || GET_AVAIL_BYTES(p) > p->keepSizeAfter) -} - - - -MY_NO_INLINE -void MatchFinder_MoveBlock(CMatchFinder *p) -{ - const size_t offset = (size_t)(p->buffer - p->bufferBase) - p->keepSizeBefore; - const size_t keepBefore = (offset & (kBlockMoveAlign - 1)) + p->keepSizeBefore; - p->buffer = p->bufferBase + keepBefore; - memmove(p->bufferBase, - p->bufferBase + (offset & ~((size_t)kBlockMoveAlign - 1)), - keepBefore + (size_t)GET_AVAIL_BYTES(p)); -} - -/* We call MoveBlock() before ReadBlock(). - So MoveBlock() can be wasteful operation, if the whole input data - can fit in current block even without calling MoveBlock(). - in important case where (dataSize <= historySize) - condition (p->blockSize > dataSize + p->keepSizeAfter) is met - So there is no MoveBlock() in that case case. -*/ - -int MatchFinder_NeedMove(CMatchFinder *p) -{ - if (p->directInput) - return 0; - if (p->streamEndWasReached || p->result != SZ_OK) - return 0; - return ((size_t)(p->bufferBase + p->blockSize - p->buffer) <= p->keepSizeAfter); -} - -void MatchFinder_ReadIfRequired(CMatchFinder *p) -{ - if (p->keepSizeAfter >= GET_AVAIL_BYTES(p)) - MatchFinder_ReadBlock(p); -} - - - -static void MatchFinder_SetDefaultSettings(CMatchFinder *p) -{ - p->cutValue = 32; - p->btMode = 1; - p->numHashBytes = 4; - p->bigHash = 0; -} - -#define kCrcPoly 0xEDB88320 - -void MatchFinder_Construct(CMatchFinder *p) -{ - unsigned i; - p->bufferBase = NULL; - p->directInput = 0; - p->hash = NULL; - p->expectedDataSize = (UInt64)(Int64)-1; - MatchFinder_SetDefaultSettings(p); - - for (i = 0; i < 256; i++) - { - UInt32 r = (UInt32)i; - unsigned j; - for (j = 0; j < 8; j++) - r = (r >> 1) ^ (kCrcPoly & ((UInt32)0 - (r & 1))); - p->crc[i] = r; - } -} - -static void MatchFinder_FreeThisClassMemory(CMatchFinder *p, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, p->hash); - p->hash = NULL; -} - -void MatchFinder_Free(CMatchFinder *p, ISzAllocPtr alloc) -{ - MatchFinder_FreeThisClassMemory(p, alloc); - LzInWindow_Free(p, alloc); -} - -static CLzRef* AllocRefs(size_t num, ISzAllocPtr alloc) -{ - size_t sizeInBytes = (size_t)num * sizeof(CLzRef); - if (sizeInBytes / sizeof(CLzRef) != num) - return NULL; - return (CLzRef *)ISzAlloc_Alloc(alloc, sizeInBytes); -} - -#if (kBlockSizeReserveMin < kBlockSizeAlign * 2) - #error Stop_Compiling_Bad_Reserve -#endif - - - -static UInt32 GetBlockSize(CMatchFinder *p, UInt32 historySize) -{ - UInt32 blockSize = (p->keepSizeBefore + p->keepSizeAfter); - /* - if (historySize > kMaxHistorySize) - return 0; - */ - // printf("\nhistorySize == 0x%x\n", historySize); - - if (p->keepSizeBefore < historySize || blockSize < p->keepSizeBefore) // if 32-bit overflow - return 0; - - { - const UInt32 kBlockSizeMax = (UInt32)0 - (UInt32)kBlockSizeAlign; - const UInt32 rem = kBlockSizeMax - blockSize; - const UInt32 reserve = (blockSize >> (blockSize < ((UInt32)1 << 30) ? 1 : 2)) - + (1 << 12) + kBlockMoveAlign + kBlockSizeAlign; // do not overflow 32-bit here - if (blockSize >= kBlockSizeMax - || rem < kBlockSizeReserveMin) // we reject settings that will be slow - return 0; - if (reserve >= rem) - blockSize = kBlockSizeMax; - else - { - blockSize += reserve; - blockSize &= ~(UInt32)(kBlockSizeAlign - 1); - } - } - // printf("\n LzFind_blockSize = %x\n", blockSize); - // printf("\n LzFind_blockSize = %d\n", blockSize >> 20); - return blockSize; -} - - -int MatchFinder_Create(CMatchFinder *p, UInt32 historySize, - UInt32 keepAddBufferBefore, UInt32 matchMaxLen, UInt32 keepAddBufferAfter, - ISzAllocPtr alloc) -{ - /* we need one additional byte in (p->keepSizeBefore), - since we use MoveBlock() after (p->pos++) and before dictionary using */ - // keepAddBufferBefore = (UInt32)0xFFFFFFFF - (1 << 22); // for debug - p->keepSizeBefore = historySize + keepAddBufferBefore + 1; - - keepAddBufferAfter += matchMaxLen; - /* we need (p->keepSizeAfter >= p->numHashBytes) */ - if (keepAddBufferAfter < p->numHashBytes) - keepAddBufferAfter = p->numHashBytes; - // keepAddBufferAfter -= 2; // for debug - p->keepSizeAfter = keepAddBufferAfter; - - if (p->directInput) - p->blockSize = 0; - if (p->directInput || LzInWindow_Create2(p, GetBlockSize(p, historySize), alloc)) - { - const UInt32 newCyclicBufferSize = historySize + 1; // do not change it - UInt32 hs; - p->matchMaxLen = matchMaxLen; - { - // UInt32 hs4; - p->fixedHashSize = 0; - hs = (1 << 16) - 1; - if (p->numHashBytes != 2) - { - hs = historySize; - if (hs > p->expectedDataSize) - hs = (UInt32)p->expectedDataSize; - if (hs != 0) - hs--; - hs |= (hs >> 1); - hs |= (hs >> 2); - hs |= (hs >> 4); - hs |= (hs >> 8); - // we propagated 16 bits in (hs). Low 16 bits must be set later - hs >>= 1; - if (hs >= (1 << 24)) - { - if (p->numHashBytes == 3) - hs = (1 << 24) - 1; - else - hs >>= 1; - /* if (bigHash) mode, GetHeads4b() in LzFindMt.c needs (hs >= ((1 << 24) - 1))) */ - } - - // hs = ((UInt32)1 << 25) - 1; // for test - - // (hash_size >= (1 << 16)) : Required for (numHashBytes > 2) - hs |= (1 << 16) - 1; /* don't change it! */ - - // bt5: we adjust the size with recommended minimum size - if (p->numHashBytes >= 5) - hs |= (256 << kLzHash_CrcShift_2) - 1; - } - p->hashMask = hs; - hs++; - - /* - hs4 = (1 << 20); - if (hs4 > hs) - hs4 = hs; - // hs4 = (1 << 16); // for test - p->hash4Mask = hs4 - 1; - */ - - if (p->numHashBytes > 2) p->fixedHashSize += kHash2Size; - if (p->numHashBytes > 3) p->fixedHashSize += kHash3Size; - // if (p->numHashBytes > 4) p->fixedHashSize += hs4; // kHash4Size; - hs += p->fixedHashSize; - } - - { - size_t newSize; - size_t numSons; - p->historySize = historySize; - p->hashSizeSum = hs; - p->cyclicBufferSize = newCyclicBufferSize; // it must be = (historySize + 1) - - numSons = newCyclicBufferSize; - if (p->btMode) - numSons <<= 1; - newSize = hs + numSons; - - // aligned size is not required here, but it can be better for some loops - #define NUM_REFS_ALIGN_MASK 0xF - newSize = (newSize + NUM_REFS_ALIGN_MASK) & ~(size_t)NUM_REFS_ALIGN_MASK; - - if (p->hash && p->numRefs == newSize) - return 1; - - MatchFinder_FreeThisClassMemory(p, alloc); - p->numRefs = newSize; - p->hash = AllocRefs(newSize, alloc); - - if (p->hash) - { - p->son = p->hash + p->hashSizeSum; - return 1; - } - } - } - - MatchFinder_Free(p, alloc); - return 0; -} - - -static void MatchFinder_SetLimits(CMatchFinder *p) -{ - UInt32 k; - UInt32 n = kMaxValForNormalize - p->pos; - if (n == 0) - n = (UInt32)(Int32)-1; // we allow (pos == 0) at start even with (kMaxValForNormalize == 0) - - k = p->cyclicBufferSize - p->cyclicBufferPos; - if (k < n) - n = k; - - k = GET_AVAIL_BYTES(p); - { - const UInt32 ksa = p->keepSizeAfter; - UInt32 mm = p->matchMaxLen; - if (k > ksa) - k -= ksa; // we must limit exactly to keepSizeAfter for ReadBlock - else if (k >= mm) - { - // the limitation for (p->lenLimit) update - k -= mm; // optimization : to reduce the number of checks - k++; - // k = 1; // non-optimized version : for debug - } - else - { - mm = k; - if (k != 0) - k = 1; - } - p->lenLimit = mm; - } - if (k < n) - n = k; - - p->posLimit = p->pos + n; -} - - -void MatchFinder_Init_LowHash(CMatchFinder *p) -{ - size_t i; - CLzRef *items = p->hash; - const size_t numItems = p->fixedHashSize; - for (i = 0; i < numItems; i++) - items[i] = kEmptyHashValue; -} - - -void MatchFinder_Init_HighHash(CMatchFinder *p) -{ - size_t i; - CLzRef *items = p->hash + p->fixedHashSize; - const size_t numItems = (size_t)p->hashMask + 1; - for (i = 0; i < numItems; i++) - items[i] = kEmptyHashValue; -} - - -void MatchFinder_Init_4(CMatchFinder *p) -{ - p->buffer = p->bufferBase; - { - /* kEmptyHashValue = 0 (Zero) is used in hash tables as NO-VALUE marker. - the code in CMatchFinderMt expects (pos = 1) */ - p->pos = - p->streamPos = - 1; // it's smallest optimal value. do not change it - // 0; // for debug - } - p->result = SZ_OK; - p->streamEndWasReached = 0; -} - - -// (CYC_TO_POS_OFFSET == 0) is expected by some optimized code -#define CYC_TO_POS_OFFSET 0 -// #define CYC_TO_POS_OFFSET 1 // for debug - -void MatchFinder_Init(CMatchFinder *p) -{ - MatchFinder_Init_HighHash(p); - MatchFinder_Init_LowHash(p); - MatchFinder_Init_4(p); - // if (readData) - MatchFinder_ReadBlock(p); - - /* if we init (cyclicBufferPos = pos), then we can use one variable - instead of both (cyclicBufferPos) and (pos) : only before (cyclicBufferPos) wrapping */ - p->cyclicBufferPos = (p->pos - CYC_TO_POS_OFFSET); // init with relation to (pos) - // p->cyclicBufferPos = 0; // smallest value - // p->son[0] = p->son[1] = 0; // unused: we can init skipped record for speculated accesses. - MatchFinder_SetLimits(p); -} - - - -#ifdef MY_CPU_X86_OR_AMD64 - #if defined(__clang__) && (__clang_major__ >= 8) \ - || defined(__GNUC__) && (__GNUC__ >= 8) \ - || defined(__INTEL_COMPILER) && (__INTEL_COMPILER >= 1900) - #define USE_SATUR_SUB_128 - #define USE_AVX2 - #define ATTRIB_SSE41 __attribute__((__target__("sse4.1"))) - #define ATTRIB_AVX2 __attribute__((__target__("avx2"))) - #elif defined(_MSC_VER) - #if (_MSC_VER >= 1600) - #define USE_SATUR_SUB_128 - #if (_MSC_VER >= 1900) - #define USE_AVX2 - #include // avx - #endif - #endif - #endif - -// #elif defined(MY_CPU_ARM_OR_ARM64) -#elif defined(MY_CPU_ARM64) - - #if defined(__clang__) && (__clang_major__ >= 8) \ - || defined(__GNUC__) && (__GNUC__ >= 8) - #define USE_SATUR_SUB_128 - #ifdef MY_CPU_ARM64 - // #define ATTRIB_SSE41 __attribute__((__target__(""))) - #else - // #define ATTRIB_SSE41 __attribute__((__target__("fpu=crypto-neon-fp-armv8"))) - #endif - - #elif defined(_MSC_VER) - #if (_MSC_VER >= 1910) - #define USE_SATUR_SUB_128 - #endif - #endif - - #if defined(_MSC_VER) && defined(MY_CPU_ARM64) - #include - #else - #include - #endif - -#endif - -/* -#ifndef ATTRIB_SSE41 - #define ATTRIB_SSE41 -#endif -#ifndef ATTRIB_AVX2 - #define ATTRIB_AVX2 -#endif -*/ - -#ifdef USE_SATUR_SUB_128 - -// #define _SHOW_HW_STATUS - -#ifdef _SHOW_HW_STATUS -#include -#define _PRF(x) x -_PRF(;) -#else -#define _PRF(x) -#endif - -#ifdef MY_CPU_ARM_OR_ARM64 - -#ifdef MY_CPU_ARM64 -// #define FORCE_SATUR_SUB_128 -#endif - -typedef uint32x4_t v128; -#define SASUB_128(i) \ - *(v128 *)(void *)(items + (i) * 4) = \ - vsubq_u32(vmaxq_u32(*(const v128 *)(const void *)(items + (i) * 4), sub2), sub2); - -#else - -#include // sse4.1 - -typedef __m128i v128; -#define SASUB_128(i) \ - *(v128 *)(void *)(items + (i) * 4) = \ - _mm_sub_epi32(_mm_max_epu32(*(const v128 *)(const void *)(items + (i) * 4), sub2), sub2); // SSE 4.1 - -#endif - - - -MY_NO_INLINE -static -#ifdef ATTRIB_SSE41 -ATTRIB_SSE41 -#endif -void -MY_FAST_CALL -LzFind_SaturSub_128(UInt32 subValue, CLzRef *items, const CLzRef *lim) -{ - v128 sub2 = - #ifdef MY_CPU_ARM_OR_ARM64 - vdupq_n_u32(subValue); - #else - _mm_set_epi32((Int32)subValue, (Int32)subValue, (Int32)subValue, (Int32)subValue); - #endif - do - { - SASUB_128(0) - SASUB_128(1) - SASUB_128(2) - SASUB_128(3) - items += 4 * 4; - } - while (items != lim); -} - - - -#ifdef USE_AVX2 - -#include // avx - -#define SASUB_256(i) *(__m256i *)(void *)(items + (i) * 8) = _mm256_sub_epi32(_mm256_max_epu32(*(const __m256i *)(const void *)(items + (i) * 8), sub2), sub2); // AVX2 - -MY_NO_INLINE -static -#ifdef ATTRIB_AVX2 -ATTRIB_AVX2 -#endif -void -MY_FAST_CALL -LzFind_SaturSub_256(UInt32 subValue, CLzRef *items, const CLzRef *lim) -{ - __m256i sub2 = _mm256_set_epi32( - (Int32)subValue, (Int32)subValue, (Int32)subValue, (Int32)subValue, - (Int32)subValue, (Int32)subValue, (Int32)subValue, (Int32)subValue); - do - { - SASUB_256(0) - SASUB_256(1) - items += 2 * 8; - } - while (items != lim); -} -#endif // USE_AVX2 - -#ifndef FORCE_SATUR_SUB_128 -typedef void (MY_FAST_CALL *LZFIND_SATUR_SUB_CODE_FUNC)( - UInt32 subValue, CLzRef *items, const CLzRef *lim); -static LZFIND_SATUR_SUB_CODE_FUNC g_LzFind_SaturSub; -#endif // FORCE_SATUR_SUB_128 - -#endif // USE_SATUR_SUB_128 - - -// kEmptyHashValue must be zero -// #define SASUB_32(i) v = items[i]; m = v - subValue; if (v < subValue) m = kEmptyHashValue; items[i] = m; -#define SASUB_32(i) v = items[i]; if (v < subValue) v = subValue; items[i] = v - subValue; - -#ifdef FORCE_SATUR_SUB_128 - -#define DEFAULT_SaturSub LzFind_SaturSub_128 - -#else - -#define DEFAULT_SaturSub LzFind_SaturSub_32 - -MY_NO_INLINE -static -void -MY_FAST_CALL -LzFind_SaturSub_32(UInt32 subValue, CLzRef *items, const CLzRef *lim) -{ - do - { - UInt32 v; - SASUB_32(0) - SASUB_32(1) - SASUB_32(2) - SASUB_32(3) - SASUB_32(4) - SASUB_32(5) - SASUB_32(6) - SASUB_32(7) - items += 8; - } - while (items != lim); -} - -#endif - - -MY_NO_INLINE -void MatchFinder_Normalize3(UInt32 subValue, CLzRef *items, size_t numItems) -{ - #define K_NORM_ALIGN_BLOCK_SIZE (1 << 6) - - CLzRef *lim; - - for (; numItems != 0 && ((unsigned)(ptrdiff_t)items & (K_NORM_ALIGN_BLOCK_SIZE - 1)) != 0; numItems--) - { - UInt32 v; - SASUB_32(0); - items++; - } - - { - #define K_NORM_ALIGN_MASK (K_NORM_ALIGN_BLOCK_SIZE / 4 - 1) - lim = items + (numItems & ~(size_t)K_NORM_ALIGN_MASK); - numItems &= K_NORM_ALIGN_MASK; - if (items != lim) - { - #if defined(USE_SATUR_SUB_128) && !defined(FORCE_SATUR_SUB_128) - if (g_LzFind_SaturSub) - g_LzFind_SaturSub(subValue, items, lim); - else - #endif - DEFAULT_SaturSub(subValue, items, lim); - } - items = lim; - } - - - for (; numItems != 0; numItems--) - { - UInt32 v; - SASUB_32(0); - items++; - } -} - - - -// call MatchFinder_CheckLimits() only after (p->pos++) update - -MY_NO_INLINE -static void MatchFinder_CheckLimits(CMatchFinder *p) -{ - if (// !p->streamEndWasReached && p->result == SZ_OK && - p->keepSizeAfter == GET_AVAIL_BYTES(p)) - { - // we try to read only in exact state (p->keepSizeAfter == GET_AVAIL_BYTES(p)) - if (MatchFinder_NeedMove(p)) - MatchFinder_MoveBlock(p); - MatchFinder_ReadBlock(p); - } - - if (p->pos == kMaxValForNormalize) - if (GET_AVAIL_BYTES(p) >= p->numHashBytes) // optional optimization for last bytes of data. - /* - if we disable normalization for last bytes of data, and - if (data_size == 4 GiB), we don't call wastfull normalization, - but (pos) will be wrapped over Zero (0) in that case. - And we cannot resume later to normal operation - */ - { - // MatchFinder_Normalize(p); - /* after normalization we need (p->pos >= p->historySize + 1); */ - /* we can reduce subValue to aligned value, if want to keep alignment - of (p->pos) and (p->buffer) for speculated accesses. */ - const UInt32 subValue = (p->pos - p->historySize - 1) /* & ~(UInt32)(kNormalizeAlign - 1) */; - // const UInt32 subValue = (1 << 15); // for debug - // printf("\nMatchFinder_Normalize() subValue == 0x%x\n", subValue); - size_t numSonRefs = p->cyclicBufferSize; - if (p->btMode) - numSonRefs <<= 1; - Inline_MatchFinder_ReduceOffsets(p, subValue); - MatchFinder_Normalize3(subValue, p->hash, (size_t)p->hashSizeSum + numSonRefs); - } - - if (p->cyclicBufferPos == p->cyclicBufferSize) - p->cyclicBufferPos = 0; - - MatchFinder_SetLimits(p); -} - - -/* - (lenLimit > maxLen) -*/ -MY_FORCE_INLINE -static UInt32 * Hc_GetMatchesSpec(size_t lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, - size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue, - UInt32 *d, unsigned maxLen) -{ - /* - son[_cyclicBufferPos] = curMatch; - for (;;) - { - UInt32 delta = pos - curMatch; - if (cutValue-- == 0 || delta >= _cyclicBufferSize) - return d; - { - const Byte *pb = cur - delta; - curMatch = son[_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)]; - if (pb[maxLen] == cur[maxLen] && *pb == *cur) - { - UInt32 len = 0; - while (++len != lenLimit) - if (pb[len] != cur[len]) - break; - if (maxLen < len) - { - maxLen = len; - *d++ = len; - *d++ = delta - 1; - if (len == lenLimit) - return d; - } - } - } - } - */ - - const Byte *lim = cur + lenLimit; - son[_cyclicBufferPos] = curMatch; - - do - { - UInt32 delta; - - if (curMatch == 0) - break; - // if (curMatch2 >= curMatch) return NULL; - delta = pos - curMatch; - if (delta >= _cyclicBufferSize) - break; - { - ptrdiff_t diff; - curMatch = son[_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)]; - diff = (ptrdiff_t)0 - (ptrdiff_t)delta; - if (cur[maxLen] == cur[(ptrdiff_t)maxLen + diff]) - { - const Byte *c = cur; - while (*c == c[diff]) - { - if (++c == lim) - { - d[0] = (UInt32)(lim - cur); - d[1] = delta - 1; - return d + 2; - } - } - { - const unsigned len = (unsigned)(c - cur); - if (maxLen < len) - { - maxLen = len; - d[0] = (UInt32)len; - d[1] = delta - 1; - d += 2; - } - } - } - } - } - while (--cutValue); - - return d; -} - - -MY_FORCE_INLINE -UInt32 * GetMatchesSpec1(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, - size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue, - UInt32 *d, UInt32 maxLen) -{ - CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; - CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); - unsigned len0 = 0, len1 = 0; - - UInt32 cmCheck; - - // if (curMatch >= pos) { *ptr0 = *ptr1 = kEmptyHashValue; return NULL; } - - cmCheck = (UInt32)(pos - _cyclicBufferSize); - if ((UInt32)pos <= _cyclicBufferSize) - cmCheck = 0; - - if (cmCheck < curMatch) - do - { - const UInt32 delta = pos - curMatch; - { - CLzRef *pair = son + ((size_t)(_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)) << 1); - const Byte *pb = cur - delta; - unsigned len = (len0 < len1 ? len0 : len1); - const UInt32 pair0 = pair[0]; - if (pb[len] == cur[len]) - { - if (++len != lenLimit && pb[len] == cur[len]) - while (++len != lenLimit) - if (pb[len] != cur[len]) - break; - if (maxLen < len) - { - maxLen = (UInt32)len; - *d++ = (UInt32)len; - *d++ = delta - 1; - if (len == lenLimit) - { - *ptr1 = pair0; - *ptr0 = pair[1]; - return d; - } - } - } - if (pb[len] < cur[len]) - { - *ptr1 = curMatch; - // const UInt32 curMatch2 = pair[1]; - // if (curMatch2 >= curMatch) { *ptr0 = *ptr1 = kEmptyHashValue; return NULL; } - // curMatch = curMatch2; - curMatch = pair[1]; - ptr1 = pair + 1; - len1 = len; - } - else - { - *ptr0 = curMatch; - curMatch = pair[0]; - ptr0 = pair; - len0 = len; - } - } - } - while(--cutValue && cmCheck < curMatch); - - *ptr0 = *ptr1 = kEmptyHashValue; - return d; -} - - -static void SkipMatchesSpec(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, - size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue) -{ - CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; - CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); - unsigned len0 = 0, len1 = 0; - - UInt32 cmCheck; - - cmCheck = (UInt32)(pos - _cyclicBufferSize); - if ((UInt32)pos <= _cyclicBufferSize) - cmCheck = 0; - - if (// curMatch >= pos || // failure - cmCheck < curMatch) - do - { - const UInt32 delta = pos - curMatch; - { - CLzRef *pair = son + ((size_t)(_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)) << 1); - const Byte *pb = cur - delta; - unsigned len = (len0 < len1 ? len0 : len1); - if (pb[len] == cur[len]) - { - while (++len != lenLimit) - if (pb[len] != cur[len]) - break; - { - if (len == lenLimit) - { - *ptr1 = pair[0]; - *ptr0 = pair[1]; - return; - } - } - } - if (pb[len] < cur[len]) - { - *ptr1 = curMatch; - curMatch = pair[1]; - ptr1 = pair + 1; - len1 = len; - } - else - { - *ptr0 = curMatch; - curMatch = pair[0]; - ptr0 = pair; - len0 = len; - } - } - } - while(--cutValue && cmCheck < curMatch); - - *ptr0 = *ptr1 = kEmptyHashValue; - return; -} - - -#define MOVE_POS \ - ++p->cyclicBufferPos; \ - p->buffer++; \ - { const UInt32 pos1 = p->pos + 1; p->pos = pos1; if (pos1 == p->posLimit) MatchFinder_CheckLimits(p); } - -#define MOVE_POS_RET MOVE_POS return distances; - -MY_NO_INLINE -static void MatchFinder_MovePos(CMatchFinder *p) -{ - /* we go here at the end of stream data, when (avail < num_hash_bytes) - We don't update sons[cyclicBufferPos << btMode]. - So (sons) record will contain junk. And we cannot resume match searching - to normal operation, even if we will provide more input data in buffer. - p->sons[p->cyclicBufferPos << p->btMode] = 0; // kEmptyHashValue - if (p->btMode) - p->sons[(p->cyclicBufferPos << p->btMode) + 1] = 0; // kEmptyHashValue - */ - MOVE_POS; -} - -#define GET_MATCHES_HEADER2(minLen, ret_op) \ - unsigned lenLimit; UInt32 hv; Byte *cur; UInt32 curMatch; \ - lenLimit = (unsigned)p->lenLimit; { if (lenLimit < minLen) { MatchFinder_MovePos(p); ret_op; }} \ - cur = p->buffer; - -#define GET_MATCHES_HEADER(minLen) GET_MATCHES_HEADER2(minLen, return distances) -#define SKIP_HEADER(minLen) do { GET_MATCHES_HEADER2(minLen, continue) - -#define MF_PARAMS(p) lenLimit, curMatch, p->pos, p->buffer, p->son, p->cyclicBufferPos, p->cyclicBufferSize, p->cutValue - -#define SKIP_FOOTER SkipMatchesSpec(MF_PARAMS(p)); MOVE_POS; } while (--num); - -#define GET_MATCHES_FOOTER_BASE(_maxLen_, func) \ - distances = func(MF_PARAMS(p), \ - distances, (UInt32)_maxLen_); MOVE_POS_RET; - -#define GET_MATCHES_FOOTER_BT(_maxLen_) \ - GET_MATCHES_FOOTER_BASE(_maxLen_, GetMatchesSpec1) - -#define GET_MATCHES_FOOTER_HC(_maxLen_) \ - GET_MATCHES_FOOTER_BASE(_maxLen_, Hc_GetMatchesSpec) - - - -#define UPDATE_maxLen { \ - const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)d2; \ - const Byte *c = cur + maxLen; \ - const Byte *lim = cur + lenLimit; \ - for (; c != lim; c++) if (*(c + diff) != *c) break; \ - maxLen = (unsigned)(c - cur); } - -static UInt32* Bt2_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) -{ - GET_MATCHES_HEADER(2) - HASH2_CALC; - curMatch = p->hash[hv]; - p->hash[hv] = p->pos; - GET_MATCHES_FOOTER_BT(1) -} - -UInt32* Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) -{ - GET_MATCHES_HEADER(3) - HASH_ZIP_CALC; - curMatch = p->hash[hv]; - p->hash[hv] = p->pos; - GET_MATCHES_FOOTER_BT(2) -} - - -#define SET_mmm \ - mmm = p->cyclicBufferSize; \ - if (pos < mmm) \ - mmm = pos; - - -static UInt32* Bt3_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) -{ - UInt32 mmm; - UInt32 h2, d2, pos; - unsigned maxLen; - UInt32 *hash; - GET_MATCHES_HEADER(3) - - HASH3_CALC; - - hash = p->hash; - pos = p->pos; - - d2 = pos - hash[h2]; - - curMatch = (hash + kFix3HashSize)[hv]; - - hash[h2] = pos; - (hash + kFix3HashSize)[hv] = pos; - - SET_mmm - - maxLen = 2; - - if (d2 < mmm && *(cur - d2) == *cur) - { - UPDATE_maxLen - distances[0] = (UInt32)maxLen; - distances[1] = d2 - 1; - distances += 2; - if (maxLen == lenLimit) - { - SkipMatchesSpec(MF_PARAMS(p)); - MOVE_POS_RET; - } - } - - GET_MATCHES_FOOTER_BT(maxLen) -} - - -static UInt32* Bt4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) -{ - UInt32 mmm; - UInt32 h2, h3, d2, d3, pos; - unsigned maxLen; - UInt32 *hash; - GET_MATCHES_HEADER(4) - - HASH4_CALC; - - hash = p->hash; - pos = p->pos; - - d2 = pos - hash [h2]; - d3 = pos - (hash + kFix3HashSize)[h3]; - curMatch = (hash + kFix4HashSize)[hv]; - - hash [h2] = pos; - (hash + kFix3HashSize)[h3] = pos; - (hash + kFix4HashSize)[hv] = pos; - - SET_mmm - - maxLen = 3; - - for (;;) - { - if (d2 < mmm && *(cur - d2) == *cur) - { - distances[0] = 2; - distances[1] = d2 - 1; - distances += 2; - if (*(cur - d2 + 2) == cur[2]) - { - // distances[-2] = 3; - } - else if (d3 < mmm && *(cur - d3) == *cur) - { - d2 = d3; - distances[1] = d3 - 1; - distances += 2; - } - else - break; - } - else if (d3 < mmm && *(cur - d3) == *cur) - { - d2 = d3; - distances[1] = d3 - 1; - distances += 2; - } - else - break; - - UPDATE_maxLen - distances[-2] = (UInt32)maxLen; - if (maxLen == lenLimit) - { - SkipMatchesSpec(MF_PARAMS(p)); - MOVE_POS_RET - } - break; - } - - GET_MATCHES_FOOTER_BT(maxLen) -} - - -static UInt32* Bt5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) -{ - UInt32 mmm; - UInt32 h2, h3, d2, d3, maxLen, pos; - UInt32 *hash; - GET_MATCHES_HEADER(5) - - HASH5_CALC; - - hash = p->hash; - pos = p->pos; - - d2 = pos - hash [h2]; - d3 = pos - (hash + kFix3HashSize)[h3]; - // d4 = pos - (hash + kFix4HashSize)[h4]; - - curMatch = (hash + kFix5HashSize)[hv]; - - hash [h2] = pos; - (hash + kFix3HashSize)[h3] = pos; - // (hash + kFix4HashSize)[h4] = pos; - (hash + kFix5HashSize)[hv] = pos; - - SET_mmm - - maxLen = 4; - - for (;;) - { - if (d2 < mmm && *(cur - d2) == *cur) - { - distances[0] = 2; - distances[1] = d2 - 1; - distances += 2; - if (*(cur - d2 + 2) == cur[2]) - { - } - else if (d3 < mmm && *(cur - d3) == *cur) - { - distances[1] = d3 - 1; - distances += 2; - d2 = d3; - } - else - break; - } - else if (d3 < mmm && *(cur - d3) == *cur) - { - distances[1] = d3 - 1; - distances += 2; - d2 = d3; - } - else - break; - - distances[-2] = 3; - if (*(cur - d2 + 3) != cur[3]) - break; - UPDATE_maxLen - distances[-2] = (UInt32)maxLen; - if (maxLen == lenLimit) - { - SkipMatchesSpec(MF_PARAMS(p)); - MOVE_POS_RET; - } - break; - } - - GET_MATCHES_FOOTER_BT(maxLen) -} - - -static UInt32* Hc4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) -{ - UInt32 mmm; - UInt32 h2, h3, d2, d3, pos; - unsigned maxLen; - UInt32 *hash; - GET_MATCHES_HEADER(4) - - HASH4_CALC; - - hash = p->hash; - pos = p->pos; - - d2 = pos - hash [h2]; - d3 = pos - (hash + kFix3HashSize)[h3]; - curMatch = (hash + kFix4HashSize)[hv]; - - hash [h2] = pos; - (hash + kFix3HashSize)[h3] = pos; - (hash + kFix4HashSize)[hv] = pos; - - SET_mmm - - maxLen = 3; - - for (;;) - { - if (d2 < mmm && *(cur - d2) == *cur) - { - distances[0] = 2; - distances[1] = d2 - 1; - distances += 2; - if (*(cur - d2 + 2) == cur[2]) - { - // distances[-2] = 3; - } - else if (d3 < mmm && *(cur - d3) == *cur) - { - d2 = d3; - distances[1] = d3 - 1; - distances += 2; - } - else - break; - } - else if (d3 < mmm && *(cur - d3) == *cur) - { - d2 = d3; - distances[1] = d3 - 1; - distances += 2; - } - else - break; - - UPDATE_maxLen - distances[-2] = (UInt32)maxLen; - if (maxLen == lenLimit) - { - p->son[p->cyclicBufferPos] = curMatch; - MOVE_POS_RET; - } - break; - } - - GET_MATCHES_FOOTER_HC(maxLen); -} - - -static UInt32 * Hc5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) -{ - UInt32 mmm; - UInt32 h2, h3, d2, d3, maxLen, pos; - UInt32 *hash; - GET_MATCHES_HEADER(5) - - HASH5_CALC; - - hash = p->hash; - pos = p->pos; - - d2 = pos - hash [h2]; - d3 = pos - (hash + kFix3HashSize)[h3]; - // d4 = pos - (hash + kFix4HashSize)[h4]; - - curMatch = (hash + kFix5HashSize)[hv]; - - hash [h2] = pos; - (hash + kFix3HashSize)[h3] = pos; - // (hash + kFix4HashSize)[h4] = pos; - (hash + kFix5HashSize)[hv] = pos; - - SET_mmm - - maxLen = 4; - - for (;;) - { - if (d2 < mmm && *(cur - d2) == *cur) - { - distances[0] = 2; - distances[1] = d2 - 1; - distances += 2; - if (*(cur - d2 + 2) == cur[2]) - { - } - else if (d3 < mmm && *(cur - d3) == *cur) - { - distances[1] = d3 - 1; - distances += 2; - d2 = d3; - } - else - break; - } - else if (d3 < mmm && *(cur - d3) == *cur) - { - distances[1] = d3 - 1; - distances += 2; - d2 = d3; - } - else - break; - - distances[-2] = 3; - if (*(cur - d2 + 3) != cur[3]) - break; - UPDATE_maxLen - distances[-2] = maxLen; - if (maxLen == lenLimit) - { - p->son[p->cyclicBufferPos] = curMatch; - MOVE_POS_RET; - } - break; - } - - GET_MATCHES_FOOTER_HC(maxLen); -} - - -UInt32* Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) -{ - GET_MATCHES_HEADER(3) - HASH_ZIP_CALC; - curMatch = p->hash[hv]; - p->hash[hv] = p->pos; - GET_MATCHES_FOOTER_HC(2) -} - - -static void Bt2_MatchFinder_Skip(CMatchFinder *p, UInt32 num) -{ - SKIP_HEADER(2) - { - HASH2_CALC; - curMatch = p->hash[hv]; - p->hash[hv] = p->pos; - } - SKIP_FOOTER -} - -void Bt3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num) -{ - SKIP_HEADER(3) - { - HASH_ZIP_CALC; - curMatch = p->hash[hv]; - p->hash[hv] = p->pos; - } - SKIP_FOOTER -} - -static void Bt3_MatchFinder_Skip(CMatchFinder *p, UInt32 num) -{ - SKIP_HEADER(3) - { - UInt32 h2; - UInt32 *hash; - HASH3_CALC; - hash = p->hash; - curMatch = (hash + kFix3HashSize)[hv]; - hash[h2] = - (hash + kFix3HashSize)[hv] = p->pos; - } - SKIP_FOOTER -} - -static void Bt4_MatchFinder_Skip(CMatchFinder *p, UInt32 num) -{ - SKIP_HEADER(4) - { - UInt32 h2, h3; - UInt32 *hash; - HASH4_CALC; - hash = p->hash; - curMatch = (hash + kFix4HashSize)[hv]; - hash [h2] = - (hash + kFix3HashSize)[h3] = - (hash + kFix4HashSize)[hv] = p->pos; - } - SKIP_FOOTER -} - -static void Bt5_MatchFinder_Skip(CMatchFinder *p, UInt32 num) -{ - SKIP_HEADER(5) - { - UInt32 h2, h3; - UInt32 *hash; - HASH5_CALC; - hash = p->hash; - curMatch = (hash + kFix5HashSize)[hv]; - hash [h2] = - (hash + kFix3HashSize)[h3] = - // (hash + kFix4HashSize)[h4] = - (hash + kFix5HashSize)[hv] = p->pos; - } - SKIP_FOOTER -} - - -#define HC_SKIP_HEADER(minLen) \ - do { if (p->lenLimit < minLen) { MatchFinder_MovePos(p); num--; continue; } { \ - Byte *cur; \ - UInt32 *hash; \ - UInt32 *son; \ - UInt32 pos = p->pos; \ - UInt32 num2 = num; \ - /* (p->pos == p->posLimit) is not allowed here !!! */ \ - { const UInt32 rem = p->posLimit - pos; if (num2 > rem) num2 = rem; } \ - num -= num2; \ - { const UInt32 cycPos = p->cyclicBufferPos; \ - son = p->son + cycPos; \ - p->cyclicBufferPos = cycPos + num2; } \ - cur = p->buffer; \ - hash = p->hash; \ - do { \ - UInt32 curMatch; \ - UInt32 hv; - - -#define HC_SKIP_FOOTER \ - cur++; pos++; *son++ = curMatch; \ - } while (--num2); \ - p->buffer = cur; \ - p->pos = pos; \ - if (pos == p->posLimit) MatchFinder_CheckLimits(p); \ - }} while(num); \ - - -static void Hc4_MatchFinder_Skip(CMatchFinder *p, UInt32 num) -{ - HC_SKIP_HEADER(4) - - UInt32 h2, h3; - HASH4_CALC; - curMatch = (hash + kFix4HashSize)[hv]; - hash [h2] = - (hash + kFix3HashSize)[h3] = - (hash + kFix4HashSize)[hv] = pos; - - HC_SKIP_FOOTER -} - - -static void Hc5_MatchFinder_Skip(CMatchFinder *p, UInt32 num) -{ - HC_SKIP_HEADER(5) - - UInt32 h2, h3; - HASH5_CALC - curMatch = (hash + kFix5HashSize)[hv]; - hash [h2] = - (hash + kFix3HashSize)[h3] = - // (hash + kFix4HashSize)[h4] = - (hash + kFix5HashSize)[hv] = pos; - - HC_SKIP_FOOTER -} - - -void Hc3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num) -{ - HC_SKIP_HEADER(3) - - HASH_ZIP_CALC; - curMatch = hash[hv]; - hash[hv] = pos; - - HC_SKIP_FOOTER -} - - -void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder2 *vTable) -{ - vTable->Init = (Mf_Init_Func)MatchFinder_Init; - vTable->GetNumAvailableBytes = (Mf_GetNumAvailableBytes_Func)MatchFinder_GetNumAvailableBytes; - vTable->GetPointerToCurrentPos = (Mf_GetPointerToCurrentPos_Func)MatchFinder_GetPointerToCurrentPos; - if (!p->btMode) - { - if (p->numHashBytes <= 4) - { - vTable->GetMatches = (Mf_GetMatches_Func)Hc4_MatchFinder_GetMatches; - vTable->Skip = (Mf_Skip_Func)Hc4_MatchFinder_Skip; - } - else - { - vTable->GetMatches = (Mf_GetMatches_Func)Hc5_MatchFinder_GetMatches; - vTable->Skip = (Mf_Skip_Func)Hc5_MatchFinder_Skip; - } - } - else if (p->numHashBytes == 2) - { - vTable->GetMatches = (Mf_GetMatches_Func)Bt2_MatchFinder_GetMatches; - vTable->Skip = (Mf_Skip_Func)Bt2_MatchFinder_Skip; - } - else if (p->numHashBytes == 3) - { - vTable->GetMatches = (Mf_GetMatches_Func)Bt3_MatchFinder_GetMatches; - vTable->Skip = (Mf_Skip_Func)Bt3_MatchFinder_Skip; - } - else if (p->numHashBytes == 4) - { - vTable->GetMatches = (Mf_GetMatches_Func)Bt4_MatchFinder_GetMatches; - vTable->Skip = (Mf_Skip_Func)Bt4_MatchFinder_Skip; - } - else - { - vTable->GetMatches = (Mf_GetMatches_Func)Bt5_MatchFinder_GetMatches; - vTable->Skip = (Mf_Skip_Func)Bt5_MatchFinder_Skip; - } -} - - - -void LzFindPrepare() -{ - #ifndef FORCE_SATUR_SUB_128 - #ifdef USE_SATUR_SUB_128 - LZFIND_SATUR_SUB_CODE_FUNC f = NULL; - #ifdef MY_CPU_ARM_OR_ARM64 - { - if (CPU_IsSupported_NEON()) - { - // #pragma message ("=== LzFind NEON") - _PRF(printf("\n=== LzFind NEON\n")); - f = LzFind_SaturSub_128; - } - // f = 0; // for debug - } - #else // MY_CPU_ARM_OR_ARM64 - if (CPU_IsSupported_SSE41()) - { - // #pragma message ("=== LzFind SSE41") - _PRF(printf("\n=== LzFind SSE41\n")); - f = LzFind_SaturSub_128; - - #ifdef USE_AVX2 - if (CPU_IsSupported_AVX2()) - { - // #pragma message ("=== LzFind AVX2") - _PRF(printf("\n=== LzFind AVX2\n")); - f = LzFind_SaturSub_256; - } - #endif - } - #endif // MY_CPU_ARM_OR_ARM64 - g_LzFind_SaturSub = f; - #endif // USE_SATUR_SUB_128 - #endif // FORCE_SATUR_SUB_128 -} +/* LzFind.c -- Match finder for LZ algorithms +2021-11-29 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include +// #include + +#include "CpuArch.h" +#include "LzFind.h" +#include "LzHash.h" + +#define kBlockMoveAlign (1 << 7) // alignment for memmove() +#define kBlockSizeAlign (1 << 16) // alignment for block allocation +#define kBlockSizeReserveMin (1 << 24) // it's 1/256 from 4 GB dictinary + +#define kEmptyHashValue 0 + +#define kMaxValForNormalize ((UInt32)0) +// #define kMaxValForNormalize ((UInt32)(1 << 20) + 0xFFF) // for debug + +// #define kNormalizeAlign (1 << 7) // alignment for speculated accesses + +#define GET_AVAIL_BYTES(p) \ + Inline_MatchFinder_GetNumAvailableBytes(p) + + +// #define kFix5HashSize (kHash2Size + kHash3Size + kHash4Size) +#define kFix5HashSize kFix4HashSize + +/* + HASH2_CALC: + if (hv) match, then cur[0] and cur[1] also match +*/ +#define HASH2_CALC hv = GetUi16(cur); + +// (crc[0 ... 255] & 0xFF) provides one-to-one correspondence to [0 ... 255] + +/* + HASH3_CALC: + if (cur[0]) and (h2) match, then cur[1] also match + if (cur[0]) and (hv) match, then cur[1] and cur[2] also match +*/ +#define HASH3_CALC { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + hv = (temp ^ ((UInt32)cur[2] << 8)) & p->hashMask; } + +#define HASH4_CALC { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + temp ^= ((UInt32)cur[2] << 8); \ + h3 = temp & (kHash3Size - 1); \ + hv = (temp ^ (p->crc[cur[3]] << kLzHash_CrcShift_1)) & p->hashMask; } + +#define HASH5_CALC { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + temp ^= ((UInt32)cur[2] << 8); \ + h3 = temp & (kHash3Size - 1); \ + temp ^= (p->crc[cur[3]] << kLzHash_CrcShift_1); \ + /* h4 = temp & p->hash4Mask; */ /* (kHash4Size - 1); */ \ + hv = (temp ^ (p->crc[cur[4]] << kLzHash_CrcShift_2)) & p->hashMask; } + +#define HASH_ZIP_CALC hv = ((cur[2] | ((UInt32)cur[0] << 8)) ^ p->crc[cur[1]]) & 0xFFFF; + + +static void LzInWindow_Free(CMatchFinder *p, ISzAllocPtr alloc) +{ + if (!p->directInput) + { + ISzAlloc_Free(alloc, p->bufferBase); + p->bufferBase = NULL; + } +} + + +static int LzInWindow_Create2(CMatchFinder *p, UInt32 blockSize, ISzAllocPtr alloc) +{ + if (blockSize == 0) + return 0; + if (!p->bufferBase || p->blockSize != blockSize) + { + // size_t blockSizeT; + LzInWindow_Free(p, alloc); + p->blockSize = blockSize; + // blockSizeT = blockSize; + + // printf("\nblockSize = 0x%x\n", blockSize); + /* + #if defined _WIN64 + // we can allocate 4GiB, but still use UInt32 for (p->blockSize) + // we use UInt32 type for (p->blockSize), because + // we don't want to wrap over 4 GiB, + // when we use (p->streamPos - p->pos) that is UInt32. + if (blockSize >= (UInt32)0 - (UInt32)kBlockSizeAlign) + { + blockSizeT = ((size_t)1 << 32); + printf("\nchanged to blockSizeT = 4GiB\n"); + } + #endif + */ + + p->bufferBase = (Byte *)ISzAlloc_Alloc(alloc, blockSize); + // printf("\nbufferBase = %p\n", p->bufferBase); + // return 0; // for debug + } + return (p->bufferBase != NULL); +} + +static const Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p) { return p->buffer; } + +static UInt32 MatchFinder_GetNumAvailableBytes(CMatchFinder *p) { return GET_AVAIL_BYTES(p); } + + +MY_NO_INLINE +static void MatchFinder_ReadBlock(CMatchFinder *p) +{ + if (p->streamEndWasReached || p->result != SZ_OK) + return; + + /* We use (p->streamPos - p->pos) value. + (p->streamPos < p->pos) is allowed. */ + + if (p->directInput) + { + UInt32 curSize = 0xFFFFFFFF - GET_AVAIL_BYTES(p); + if (curSize > p->directInputRem) + curSize = (UInt32)p->directInputRem; + p->directInputRem -= curSize; + p->streamPos += curSize; + if (p->directInputRem == 0) + p->streamEndWasReached = 1; + return; + } + + for (;;) + { + Byte *dest = p->buffer + GET_AVAIL_BYTES(p); + size_t size = (size_t)(p->bufferBase + p->blockSize - dest); + if (size == 0) + { + /* we call ReadBlock() after NeedMove() and MoveBlock(). + NeedMove() and MoveBlock() povide more than (keepSizeAfter) + to the end of (blockSize). + So we don't execute this branch in normal code flow. + We can go here, if we will call ReadBlock() before NeedMove(), MoveBlock(). + */ + // p->result = SZ_ERROR_FAIL; // we can show error here + return; + } + + // #define kRead 3 + // if (size > kRead) size = kRead; // for debug + + p->result = ISeqInStream_Read(p->stream, dest, &size); + if (p->result != SZ_OK) + return; + if (size == 0) + { + p->streamEndWasReached = 1; + return; + } + p->streamPos += (UInt32)size; + if (GET_AVAIL_BYTES(p) > p->keepSizeAfter) + return; + /* here and in another (p->keepSizeAfter) checks we keep on 1 byte more than was requested by Create() function + (GET_AVAIL_BYTES(p) >= p->keepSizeAfter) - minimal required size */ + } + + // on exit: (p->result != SZ_OK || p->streamEndWasReached || GET_AVAIL_BYTES(p) > p->keepSizeAfter) +} + + + +MY_NO_INLINE +void MatchFinder_MoveBlock(CMatchFinder *p) +{ + const size_t offset = (size_t)(p->buffer - p->bufferBase) - p->keepSizeBefore; + const size_t keepBefore = (offset & (kBlockMoveAlign - 1)) + p->keepSizeBefore; + p->buffer = p->bufferBase + keepBefore; + memmove(p->bufferBase, + p->bufferBase + (offset & ~((size_t)kBlockMoveAlign - 1)), + keepBefore + (size_t)GET_AVAIL_BYTES(p)); +} + +/* We call MoveBlock() before ReadBlock(). + So MoveBlock() can be wasteful operation, if the whole input data + can fit in current block even without calling MoveBlock(). + in important case where (dataSize <= historySize) + condition (p->blockSize > dataSize + p->keepSizeAfter) is met + So there is no MoveBlock() in that case case. +*/ + +int MatchFinder_NeedMove(CMatchFinder *p) +{ + if (p->directInput) + return 0; + if (p->streamEndWasReached || p->result != SZ_OK) + return 0; + return ((size_t)(p->bufferBase + p->blockSize - p->buffer) <= p->keepSizeAfter); +} + +void MatchFinder_ReadIfRequired(CMatchFinder *p) +{ + if (p->keepSizeAfter >= GET_AVAIL_BYTES(p)) + MatchFinder_ReadBlock(p); +} + + + +static void MatchFinder_SetDefaultSettings(CMatchFinder *p) +{ + p->cutValue = 32; + p->btMode = 1; + p->numHashBytes = 4; + p->bigHash = 0; +} + +#define kCrcPoly 0xEDB88320 + +void MatchFinder_Construct(CMatchFinder *p) +{ + unsigned i; + p->bufferBase = NULL; + p->directInput = 0; + p->hash = NULL; + p->expectedDataSize = (UInt64)(Int64)-1; + MatchFinder_SetDefaultSettings(p); + + for (i = 0; i < 256; i++) + { + UInt32 r = (UInt32)i; + unsigned j; + for (j = 0; j < 8; j++) + r = (r >> 1) ^ (kCrcPoly & ((UInt32)0 - (r & 1))); + p->crc[i] = r; + } +} + +static void MatchFinder_FreeThisClassMemory(CMatchFinder *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->hash); + p->hash = NULL; +} + +void MatchFinder_Free(CMatchFinder *p, ISzAllocPtr alloc) +{ + MatchFinder_FreeThisClassMemory(p, alloc); + LzInWindow_Free(p, alloc); +} + +static CLzRef* AllocRefs(size_t num, ISzAllocPtr alloc) +{ + size_t sizeInBytes = (size_t)num * sizeof(CLzRef); + if (sizeInBytes / sizeof(CLzRef) != num) + return NULL; + return (CLzRef *)ISzAlloc_Alloc(alloc, sizeInBytes); +} + +#if (kBlockSizeReserveMin < kBlockSizeAlign * 2) + #error Stop_Compiling_Bad_Reserve +#endif + + + +static UInt32 GetBlockSize(CMatchFinder *p, UInt32 historySize) +{ + UInt32 blockSize = (p->keepSizeBefore + p->keepSizeAfter); + /* + if (historySize > kMaxHistorySize) + return 0; + */ + // printf("\nhistorySize == 0x%x\n", historySize); + + if (p->keepSizeBefore < historySize || blockSize < p->keepSizeBefore) // if 32-bit overflow + return 0; + + { + const UInt32 kBlockSizeMax = (UInt32)0 - (UInt32)kBlockSizeAlign; + const UInt32 rem = kBlockSizeMax - blockSize; + const UInt32 reserve = (blockSize >> (blockSize < ((UInt32)1 << 30) ? 1 : 2)) + + (1 << 12) + kBlockMoveAlign + kBlockSizeAlign; // do not overflow 32-bit here + if (blockSize >= kBlockSizeMax + || rem < kBlockSizeReserveMin) // we reject settings that will be slow + return 0; + if (reserve >= rem) + blockSize = kBlockSizeMax; + else + { + blockSize += reserve; + blockSize &= ~(UInt32)(kBlockSizeAlign - 1); + } + } + // printf("\n LzFind_blockSize = %x\n", blockSize); + // printf("\n LzFind_blockSize = %d\n", blockSize >> 20); + return blockSize; +} + + +int MatchFinder_Create(CMatchFinder *p, UInt32 historySize, + UInt32 keepAddBufferBefore, UInt32 matchMaxLen, UInt32 keepAddBufferAfter, + ISzAllocPtr alloc) +{ + /* we need one additional byte in (p->keepSizeBefore), + since we use MoveBlock() after (p->pos++) and before dictionary using */ + // keepAddBufferBefore = (UInt32)0xFFFFFFFF - (1 << 22); // for debug + p->keepSizeBefore = historySize + keepAddBufferBefore + 1; + + keepAddBufferAfter += matchMaxLen; + /* we need (p->keepSizeAfter >= p->numHashBytes) */ + if (keepAddBufferAfter < p->numHashBytes) + keepAddBufferAfter = p->numHashBytes; + // keepAddBufferAfter -= 2; // for debug + p->keepSizeAfter = keepAddBufferAfter; + + if (p->directInput) + p->blockSize = 0; + if (p->directInput || LzInWindow_Create2(p, GetBlockSize(p, historySize), alloc)) + { + const UInt32 newCyclicBufferSize = historySize + 1; // do not change it + UInt32 hs; + p->matchMaxLen = matchMaxLen; + { + // UInt32 hs4; + p->fixedHashSize = 0; + hs = (1 << 16) - 1; + if (p->numHashBytes != 2) + { + hs = historySize; + if (hs > p->expectedDataSize) + hs = (UInt32)p->expectedDataSize; + if (hs != 0) + hs--; + hs |= (hs >> 1); + hs |= (hs >> 2); + hs |= (hs >> 4); + hs |= (hs >> 8); + // we propagated 16 bits in (hs). Low 16 bits must be set later + hs >>= 1; + if (hs >= (1 << 24)) + { + if (p->numHashBytes == 3) + hs = (1 << 24) - 1; + else + hs >>= 1; + /* if (bigHash) mode, GetHeads4b() in LzFindMt.c needs (hs >= ((1 << 24) - 1))) */ + } + + // hs = ((UInt32)1 << 25) - 1; // for test + + // (hash_size >= (1 << 16)) : Required for (numHashBytes > 2) + hs |= (1 << 16) - 1; /* don't change it! */ + + // bt5: we adjust the size with recommended minimum size + if (p->numHashBytes >= 5) + hs |= (256 << kLzHash_CrcShift_2) - 1; + } + p->hashMask = hs; + hs++; + + /* + hs4 = (1 << 20); + if (hs4 > hs) + hs4 = hs; + // hs4 = (1 << 16); // for test + p->hash4Mask = hs4 - 1; + */ + + if (p->numHashBytes > 2) p->fixedHashSize += kHash2Size; + if (p->numHashBytes > 3) p->fixedHashSize += kHash3Size; + // if (p->numHashBytes > 4) p->fixedHashSize += hs4; // kHash4Size; + hs += p->fixedHashSize; + } + + { + size_t newSize; + size_t numSons; + p->historySize = historySize; + p->hashSizeSum = hs; + p->cyclicBufferSize = newCyclicBufferSize; // it must be = (historySize + 1) + + numSons = newCyclicBufferSize; + if (p->btMode) + numSons <<= 1; + newSize = hs + numSons; + + // aligned size is not required here, but it can be better for some loops + #define NUM_REFS_ALIGN_MASK 0xF + newSize = (newSize + NUM_REFS_ALIGN_MASK) & ~(size_t)NUM_REFS_ALIGN_MASK; + + if (p->hash && p->numRefs == newSize) + return 1; + + MatchFinder_FreeThisClassMemory(p, alloc); + p->numRefs = newSize; + p->hash = AllocRefs(newSize, alloc); + + if (p->hash) + { + p->son = p->hash + p->hashSizeSum; + return 1; + } + } + } + + MatchFinder_Free(p, alloc); + return 0; +} + + +static void MatchFinder_SetLimits(CMatchFinder *p) +{ + UInt32 k; + UInt32 n = kMaxValForNormalize - p->pos; + if (n == 0) + n = (UInt32)(Int32)-1; // we allow (pos == 0) at start even with (kMaxValForNormalize == 0) + + k = p->cyclicBufferSize - p->cyclicBufferPos; + if (k < n) + n = k; + + k = GET_AVAIL_BYTES(p); + { + const UInt32 ksa = p->keepSizeAfter; + UInt32 mm = p->matchMaxLen; + if (k > ksa) + k -= ksa; // we must limit exactly to keepSizeAfter for ReadBlock + else if (k >= mm) + { + // the limitation for (p->lenLimit) update + k -= mm; // optimization : to reduce the number of checks + k++; + // k = 1; // non-optimized version : for debug + } + else + { + mm = k; + if (k != 0) + k = 1; + } + p->lenLimit = mm; + } + if (k < n) + n = k; + + p->posLimit = p->pos + n; +} + + +void MatchFinder_Init_LowHash(CMatchFinder *p) +{ + size_t i; + CLzRef *items = p->hash; + const size_t numItems = p->fixedHashSize; + for (i = 0; i < numItems; i++) + items[i] = kEmptyHashValue; +} + + +void MatchFinder_Init_HighHash(CMatchFinder *p) +{ + size_t i; + CLzRef *items = p->hash + p->fixedHashSize; + const size_t numItems = (size_t)p->hashMask + 1; + for (i = 0; i < numItems; i++) + items[i] = kEmptyHashValue; +} + + +void MatchFinder_Init_4(CMatchFinder *p) +{ + p->buffer = p->bufferBase; + { + /* kEmptyHashValue = 0 (Zero) is used in hash tables as NO-VALUE marker. + the code in CMatchFinderMt expects (pos = 1) */ + p->pos = + p->streamPos = + 1; // it's smallest optimal value. do not change it + // 0; // for debug + } + p->result = SZ_OK; + p->streamEndWasReached = 0; +} + + +// (CYC_TO_POS_OFFSET == 0) is expected by some optimized code +#define CYC_TO_POS_OFFSET 0 +// #define CYC_TO_POS_OFFSET 1 // for debug + +void MatchFinder_Init(CMatchFinder *p) +{ + MatchFinder_Init_HighHash(p); + MatchFinder_Init_LowHash(p); + MatchFinder_Init_4(p); + // if (readData) + MatchFinder_ReadBlock(p); + + /* if we init (cyclicBufferPos = pos), then we can use one variable + instead of both (cyclicBufferPos) and (pos) : only before (cyclicBufferPos) wrapping */ + p->cyclicBufferPos = (p->pos - CYC_TO_POS_OFFSET); // init with relation to (pos) + // p->cyclicBufferPos = 0; // smallest value + // p->son[0] = p->son[1] = 0; // unused: we can init skipped record for speculated accesses. + MatchFinder_SetLimits(p); +} + + + +#ifdef MY_CPU_X86_OR_AMD64 + #if defined(__clang__) && (__clang_major__ >= 8) \ + || defined(__GNUC__) && (__GNUC__ >= 8) \ + || defined(__INTEL_COMPILER) && (__INTEL_COMPILER >= 1900) + #define USE_SATUR_SUB_128 + #define USE_AVX2 + #define ATTRIB_SSE41 __attribute__((__target__("sse4.1"))) + #define ATTRIB_AVX2 __attribute__((__target__("avx2"))) + #elif defined(_MSC_VER) + #if (_MSC_VER >= 1600) + #define USE_SATUR_SUB_128 + #if (_MSC_VER >= 1900) + #define USE_AVX2 + #include // avx + #endif + #endif + #endif + +// #elif defined(MY_CPU_ARM_OR_ARM64) +#elif defined(MY_CPU_ARM64) + + #if defined(__clang__) && (__clang_major__ >= 8) \ + || defined(__GNUC__) && (__GNUC__ >= 8) + #define USE_SATUR_SUB_128 + #ifdef MY_CPU_ARM64 + // #define ATTRIB_SSE41 __attribute__((__target__(""))) + #else + // #define ATTRIB_SSE41 __attribute__((__target__("fpu=crypto-neon-fp-armv8"))) + #endif + + #elif defined(_MSC_VER) + #if (_MSC_VER >= 1910) + #define USE_SATUR_SUB_128 + #endif + #endif + + #if defined(_MSC_VER) && defined(MY_CPU_ARM64) + #include + #else + #include + #endif + +#endif + +/* +#ifndef ATTRIB_SSE41 + #define ATTRIB_SSE41 +#endif +#ifndef ATTRIB_AVX2 + #define ATTRIB_AVX2 +#endif +*/ + +#ifdef USE_SATUR_SUB_128 + +// #define _SHOW_HW_STATUS + +#ifdef _SHOW_HW_STATUS +#include +#define _PRF(x) x +_PRF(;) +#else +#define _PRF(x) +#endif + +#ifdef MY_CPU_ARM_OR_ARM64 + +#ifdef MY_CPU_ARM64 +// #define FORCE_SATUR_SUB_128 +#endif + +typedef uint32x4_t v128; +#define SASUB_128(i) \ + *(v128 *)(void *)(items + (i) * 4) = \ + vsubq_u32(vmaxq_u32(*(const v128 *)(const void *)(items + (i) * 4), sub2), sub2); + +#else + +#include // sse4.1 + +typedef __m128i v128; +#define SASUB_128(i) \ + *(v128 *)(void *)(items + (i) * 4) = \ + _mm_sub_epi32(_mm_max_epu32(*(const v128 *)(const void *)(items + (i) * 4), sub2), sub2); // SSE 4.1 + +#endif + + + +MY_NO_INLINE +static +#ifdef ATTRIB_SSE41 +ATTRIB_SSE41 +#endif +void +MY_FAST_CALL +LzFind_SaturSub_128(UInt32 subValue, CLzRef *items, const CLzRef *lim) +{ + v128 sub2 = + #ifdef MY_CPU_ARM_OR_ARM64 + vdupq_n_u32(subValue); + #else + _mm_set_epi32((Int32)subValue, (Int32)subValue, (Int32)subValue, (Int32)subValue); + #endif + do + { + SASUB_128(0) + SASUB_128(1) + SASUB_128(2) + SASUB_128(3) + items += 4 * 4; + } + while (items != lim); +} + + + +#ifdef USE_AVX2 + +#include // avx + +#define SASUB_256(i) *(__m256i *)(void *)(items + (i) * 8) = _mm256_sub_epi32(_mm256_max_epu32(*(const __m256i *)(const void *)(items + (i) * 8), sub2), sub2); // AVX2 + +MY_NO_INLINE +static +#ifdef ATTRIB_AVX2 +ATTRIB_AVX2 +#endif +void +MY_FAST_CALL +LzFind_SaturSub_256(UInt32 subValue, CLzRef *items, const CLzRef *lim) +{ + __m256i sub2 = _mm256_set_epi32( + (Int32)subValue, (Int32)subValue, (Int32)subValue, (Int32)subValue, + (Int32)subValue, (Int32)subValue, (Int32)subValue, (Int32)subValue); + do + { + SASUB_256(0) + SASUB_256(1) + items += 2 * 8; + } + while (items != lim); +} +#endif // USE_AVX2 + +#ifndef FORCE_SATUR_SUB_128 +typedef void (MY_FAST_CALL *LZFIND_SATUR_SUB_CODE_FUNC)( + UInt32 subValue, CLzRef *items, const CLzRef *lim); +static LZFIND_SATUR_SUB_CODE_FUNC g_LzFind_SaturSub; +#endif // FORCE_SATUR_SUB_128 + +#endif // USE_SATUR_SUB_128 + + +// kEmptyHashValue must be zero +// #define SASUB_32(i) v = items[i]; m = v - subValue; if (v < subValue) m = kEmptyHashValue; items[i] = m; +#define SASUB_32(i) v = items[i]; if (v < subValue) v = subValue; items[i] = v - subValue; + +#ifdef FORCE_SATUR_SUB_128 + +#define DEFAULT_SaturSub LzFind_SaturSub_128 + +#else + +#define DEFAULT_SaturSub LzFind_SaturSub_32 + +MY_NO_INLINE +static +void +MY_FAST_CALL +LzFind_SaturSub_32(UInt32 subValue, CLzRef *items, const CLzRef *lim) +{ + do + { + UInt32 v; + SASUB_32(0) + SASUB_32(1) + SASUB_32(2) + SASUB_32(3) + SASUB_32(4) + SASUB_32(5) + SASUB_32(6) + SASUB_32(7) + items += 8; + } + while (items != lim); +} + +#endif + + +MY_NO_INLINE +void MatchFinder_Normalize3(UInt32 subValue, CLzRef *items, size_t numItems) +{ + #define K_NORM_ALIGN_BLOCK_SIZE (1 << 6) + + CLzRef *lim; + + for (; numItems != 0 && ((unsigned)(ptrdiff_t)items & (K_NORM_ALIGN_BLOCK_SIZE - 1)) != 0; numItems--) + { + UInt32 v; + SASUB_32(0); + items++; + } + + { + #define K_NORM_ALIGN_MASK (K_NORM_ALIGN_BLOCK_SIZE / 4 - 1) + lim = items + (numItems & ~(size_t)K_NORM_ALIGN_MASK); + numItems &= K_NORM_ALIGN_MASK; + if (items != lim) + { + #if defined(USE_SATUR_SUB_128) && !defined(FORCE_SATUR_SUB_128) + if (g_LzFind_SaturSub) + g_LzFind_SaturSub(subValue, items, lim); + else + #endif + DEFAULT_SaturSub(subValue, items, lim); + } + items = lim; + } + + + for (; numItems != 0; numItems--) + { + UInt32 v; + SASUB_32(0); + items++; + } +} + + + +// call MatchFinder_CheckLimits() only after (p->pos++) update + +MY_NO_INLINE +static void MatchFinder_CheckLimits(CMatchFinder *p) +{ + if (// !p->streamEndWasReached && p->result == SZ_OK && + p->keepSizeAfter == GET_AVAIL_BYTES(p)) + { + // we try to read only in exact state (p->keepSizeAfter == GET_AVAIL_BYTES(p)) + if (MatchFinder_NeedMove(p)) + MatchFinder_MoveBlock(p); + MatchFinder_ReadBlock(p); + } + + if (p->pos == kMaxValForNormalize) + if (GET_AVAIL_BYTES(p) >= p->numHashBytes) // optional optimization for last bytes of data. + /* + if we disable normalization for last bytes of data, and + if (data_size == 4 GiB), we don't call wastfull normalization, + but (pos) will be wrapped over Zero (0) in that case. + And we cannot resume later to normal operation + */ + { + // MatchFinder_Normalize(p); + /* after normalization we need (p->pos >= p->historySize + 1); */ + /* we can reduce subValue to aligned value, if want to keep alignment + of (p->pos) and (p->buffer) for speculated accesses. */ + const UInt32 subValue = (p->pos - p->historySize - 1) /* & ~(UInt32)(kNormalizeAlign - 1) */; + // const UInt32 subValue = (1 << 15); // for debug + // printf("\nMatchFinder_Normalize() subValue == 0x%x\n", subValue); + size_t numSonRefs = p->cyclicBufferSize; + if (p->btMode) + numSonRefs <<= 1; + Inline_MatchFinder_ReduceOffsets(p, subValue); + MatchFinder_Normalize3(subValue, p->hash, (size_t)p->hashSizeSum + numSonRefs); + } + + if (p->cyclicBufferPos == p->cyclicBufferSize) + p->cyclicBufferPos = 0; + + MatchFinder_SetLimits(p); +} + + +/* + (lenLimit > maxLen) +*/ +MY_FORCE_INLINE +static UInt32 * Hc_GetMatchesSpec(size_t lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue, + UInt32 *d, unsigned maxLen) +{ + /* + son[_cyclicBufferPos] = curMatch; + for (;;) + { + UInt32 delta = pos - curMatch; + if (cutValue-- == 0 || delta >= _cyclicBufferSize) + return d; + { + const Byte *pb = cur - delta; + curMatch = son[_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)]; + if (pb[maxLen] == cur[maxLen] && *pb == *cur) + { + UInt32 len = 0; + while (++len != lenLimit) + if (pb[len] != cur[len]) + break; + if (maxLen < len) + { + maxLen = len; + *d++ = len; + *d++ = delta - 1; + if (len == lenLimit) + return d; + } + } + } + } + */ + + const Byte *lim = cur + lenLimit; + son[_cyclicBufferPos] = curMatch; + + do + { + UInt32 delta; + + if (curMatch == 0) + break; + // if (curMatch2 >= curMatch) return NULL; + delta = pos - curMatch; + if (delta >= _cyclicBufferSize) + break; + { + ptrdiff_t diff; + curMatch = son[_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)]; + diff = (ptrdiff_t)0 - (ptrdiff_t)delta; + if (cur[maxLen] == cur[(ptrdiff_t)maxLen + diff]) + { + const Byte *c = cur; + while (*c == c[diff]) + { + if (++c == lim) + { + d[0] = (UInt32)(lim - cur); + d[1] = delta - 1; + return d + 2; + } + } + { + const unsigned len = (unsigned)(c - cur); + if (maxLen < len) + { + maxLen = len; + d[0] = (UInt32)len; + d[1] = delta - 1; + d += 2; + } + } + } + } + } + while (--cutValue); + + return d; +} + + +MY_FORCE_INLINE +UInt32 * GetMatchesSpec1(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue, + UInt32 *d, UInt32 maxLen) +{ + CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + unsigned len0 = 0, len1 = 0; + + UInt32 cmCheck; + + // if (curMatch >= pos) { *ptr0 = *ptr1 = kEmptyHashValue; return NULL; } + + cmCheck = (UInt32)(pos - _cyclicBufferSize); + if ((UInt32)pos <= _cyclicBufferSize) + cmCheck = 0; + + if (cmCheck < curMatch) + do + { + const UInt32 delta = pos - curMatch; + { + CLzRef *pair = son + ((size_t)(_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)) << 1); + const Byte *pb = cur - delta; + unsigned len = (len0 < len1 ? len0 : len1); + const UInt32 pair0 = pair[0]; + if (pb[len] == cur[len]) + { + if (++len != lenLimit && pb[len] == cur[len]) + while (++len != lenLimit) + if (pb[len] != cur[len]) + break; + if (maxLen < len) + { + maxLen = (UInt32)len; + *d++ = (UInt32)len; + *d++ = delta - 1; + if (len == lenLimit) + { + *ptr1 = pair0; + *ptr0 = pair[1]; + return d; + } + } + } + if (pb[len] < cur[len]) + { + *ptr1 = curMatch; + // const UInt32 curMatch2 = pair[1]; + // if (curMatch2 >= curMatch) { *ptr0 = *ptr1 = kEmptyHashValue; return NULL; } + // curMatch = curMatch2; + curMatch = pair[1]; + ptr1 = pair + 1; + len1 = len; + } + else + { + *ptr0 = curMatch; + curMatch = pair[0]; + ptr0 = pair; + len0 = len; + } + } + } + while(--cutValue && cmCheck < curMatch); + + *ptr0 = *ptr1 = kEmptyHashValue; + return d; +} + + +static void SkipMatchesSpec(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue) +{ + CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + unsigned len0 = 0, len1 = 0; + + UInt32 cmCheck; + + cmCheck = (UInt32)(pos - _cyclicBufferSize); + if ((UInt32)pos <= _cyclicBufferSize) + cmCheck = 0; + + if (// curMatch >= pos || // failure + cmCheck < curMatch) + do + { + const UInt32 delta = pos - curMatch; + { + CLzRef *pair = son + ((size_t)(_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)) << 1); + const Byte *pb = cur - delta; + unsigned len = (len0 < len1 ? len0 : len1); + if (pb[len] == cur[len]) + { + while (++len != lenLimit) + if (pb[len] != cur[len]) + break; + { + if (len == lenLimit) + { + *ptr1 = pair[0]; + *ptr0 = pair[1]; + return; + } + } + } + if (pb[len] < cur[len]) + { + *ptr1 = curMatch; + curMatch = pair[1]; + ptr1 = pair + 1; + len1 = len; + } + else + { + *ptr0 = curMatch; + curMatch = pair[0]; + ptr0 = pair; + len0 = len; + } + } + } + while(--cutValue && cmCheck < curMatch); + + *ptr0 = *ptr1 = kEmptyHashValue; + return; +} + + +#define MOVE_POS \ + ++p->cyclicBufferPos; \ + p->buffer++; \ + { const UInt32 pos1 = p->pos + 1; p->pos = pos1; if (pos1 == p->posLimit) MatchFinder_CheckLimits(p); } + +#define MOVE_POS_RET MOVE_POS return distances; + +MY_NO_INLINE +static void MatchFinder_MovePos(CMatchFinder *p) +{ + /* we go here at the end of stream data, when (avail < num_hash_bytes) + We don't update sons[cyclicBufferPos << btMode]. + So (sons) record will contain junk. And we cannot resume match searching + to normal operation, even if we will provide more input data in buffer. + p->sons[p->cyclicBufferPos << p->btMode] = 0; // kEmptyHashValue + if (p->btMode) + p->sons[(p->cyclicBufferPos << p->btMode) + 1] = 0; // kEmptyHashValue + */ + MOVE_POS; +} + +#define GET_MATCHES_HEADER2(minLen, ret_op) \ + unsigned lenLimit; UInt32 hv; Byte *cur; UInt32 curMatch; \ + lenLimit = (unsigned)p->lenLimit; { if (lenLimit < minLen) { MatchFinder_MovePos(p); ret_op; }} \ + cur = p->buffer; + +#define GET_MATCHES_HEADER(minLen) GET_MATCHES_HEADER2(minLen, return distances) +#define SKIP_HEADER(minLen) do { GET_MATCHES_HEADER2(minLen, continue) + +#define MF_PARAMS(p) lenLimit, curMatch, p->pos, p->buffer, p->son, p->cyclicBufferPos, p->cyclicBufferSize, p->cutValue + +#define SKIP_FOOTER SkipMatchesSpec(MF_PARAMS(p)); MOVE_POS; } while (--num); + +#define GET_MATCHES_FOOTER_BASE(_maxLen_, func) \ + distances = func(MF_PARAMS(p), \ + distances, (UInt32)_maxLen_); MOVE_POS_RET; + +#define GET_MATCHES_FOOTER_BT(_maxLen_) \ + GET_MATCHES_FOOTER_BASE(_maxLen_, GetMatchesSpec1) + +#define GET_MATCHES_FOOTER_HC(_maxLen_) \ + GET_MATCHES_FOOTER_BASE(_maxLen_, Hc_GetMatchesSpec) + + + +#define UPDATE_maxLen { \ + const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)d2; \ + const Byte *c = cur + maxLen; \ + const Byte *lim = cur + lenLimit; \ + for (; c != lim; c++) if (*(c + diff) != *c) break; \ + maxLen = (unsigned)(c - cur); } + +static UInt32* Bt2_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + GET_MATCHES_HEADER(2) + HASH2_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + GET_MATCHES_FOOTER_BT(1) +} + +UInt32* Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + GET_MATCHES_HEADER(3) + HASH_ZIP_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + GET_MATCHES_FOOTER_BT(2) +} + + +#define SET_mmm \ + mmm = p->cyclicBufferSize; \ + if (pos < mmm) \ + mmm = pos; + + +static UInt32* Bt3_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 mmm; + UInt32 h2, d2, pos; + unsigned maxLen; + UInt32 *hash; + GET_MATCHES_HEADER(3) + + HASH3_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash[h2]; + + curMatch = (hash + kFix3HashSize)[hv]; + + hash[h2] = pos; + (hash + kFix3HashSize)[hv] = pos; + + SET_mmm + + maxLen = 2; + + if (d2 < mmm && *(cur - d2) == *cur) + { + UPDATE_maxLen + distances[0] = (UInt32)maxLen; + distances[1] = d2 - 1; + distances += 2; + if (maxLen == lenLimit) + { + SkipMatchesSpec(MF_PARAMS(p)); + MOVE_POS_RET; + } + } + + GET_MATCHES_FOOTER_BT(maxLen) +} + + +static UInt32* Bt4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 mmm; + UInt32 h2, h3, d2, d3, pos; + unsigned maxLen; + UInt32 *hash; + GET_MATCHES_HEADER(4) + + HASH4_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash [h2]; + d3 = pos - (hash + kFix3HashSize)[h3]; + curMatch = (hash + kFix4HashSize)[hv]; + + hash [h2] = pos; + (hash + kFix3HashSize)[h3] = pos; + (hash + kFix4HashSize)[hv] = pos; + + SET_mmm + + maxLen = 3; + + for (;;) + { + if (d2 < mmm && *(cur - d2) == *cur) + { + distances[0] = 2; + distances[1] = d2 - 1; + distances += 2; + if (*(cur - d2 + 2) == cur[2]) + { + // distances[-2] = 3; + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + d2 = d3; + distances[1] = d3 - 1; + distances += 2; + } + else + break; + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + d2 = d3; + distances[1] = d3 - 1; + distances += 2; + } + else + break; + + UPDATE_maxLen + distances[-2] = (UInt32)maxLen; + if (maxLen == lenLimit) + { + SkipMatchesSpec(MF_PARAMS(p)); + MOVE_POS_RET + } + break; + } + + GET_MATCHES_FOOTER_BT(maxLen) +} + + +static UInt32* Bt5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 mmm; + UInt32 h2, h3, d2, d3, maxLen, pos; + UInt32 *hash; + GET_MATCHES_HEADER(5) + + HASH5_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash [h2]; + d3 = pos - (hash + kFix3HashSize)[h3]; + // d4 = pos - (hash + kFix4HashSize)[h4]; + + curMatch = (hash + kFix5HashSize)[hv]; + + hash [h2] = pos; + (hash + kFix3HashSize)[h3] = pos; + // (hash + kFix4HashSize)[h4] = pos; + (hash + kFix5HashSize)[hv] = pos; + + SET_mmm + + maxLen = 4; + + for (;;) + { + if (d2 < mmm && *(cur - d2) == *cur) + { + distances[0] = 2; + distances[1] = d2 - 1; + distances += 2; + if (*(cur - d2 + 2) == cur[2]) + { + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + distances[1] = d3 - 1; + distances += 2; + d2 = d3; + } + else + break; + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + distances[1] = d3 - 1; + distances += 2; + d2 = d3; + } + else + break; + + distances[-2] = 3; + if (*(cur - d2 + 3) != cur[3]) + break; + UPDATE_maxLen + distances[-2] = (UInt32)maxLen; + if (maxLen == lenLimit) + { + SkipMatchesSpec(MF_PARAMS(p)); + MOVE_POS_RET; + } + break; + } + + GET_MATCHES_FOOTER_BT(maxLen) +} + + +static UInt32* Hc4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 mmm; + UInt32 h2, h3, d2, d3, pos; + unsigned maxLen; + UInt32 *hash; + GET_MATCHES_HEADER(4) + + HASH4_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash [h2]; + d3 = pos - (hash + kFix3HashSize)[h3]; + curMatch = (hash + kFix4HashSize)[hv]; + + hash [h2] = pos; + (hash + kFix3HashSize)[h3] = pos; + (hash + kFix4HashSize)[hv] = pos; + + SET_mmm + + maxLen = 3; + + for (;;) + { + if (d2 < mmm && *(cur - d2) == *cur) + { + distances[0] = 2; + distances[1] = d2 - 1; + distances += 2; + if (*(cur - d2 + 2) == cur[2]) + { + // distances[-2] = 3; + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + d2 = d3; + distances[1] = d3 - 1; + distances += 2; + } + else + break; + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + d2 = d3; + distances[1] = d3 - 1; + distances += 2; + } + else + break; + + UPDATE_maxLen + distances[-2] = (UInt32)maxLen; + if (maxLen == lenLimit) + { + p->son[p->cyclicBufferPos] = curMatch; + MOVE_POS_RET; + } + break; + } + + GET_MATCHES_FOOTER_HC(maxLen); +} + + +static UInt32 * Hc5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 mmm; + UInt32 h2, h3, d2, d3, maxLen, pos; + UInt32 *hash; + GET_MATCHES_HEADER(5) + + HASH5_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash [h2]; + d3 = pos - (hash + kFix3HashSize)[h3]; + // d4 = pos - (hash + kFix4HashSize)[h4]; + + curMatch = (hash + kFix5HashSize)[hv]; + + hash [h2] = pos; + (hash + kFix3HashSize)[h3] = pos; + // (hash + kFix4HashSize)[h4] = pos; + (hash + kFix5HashSize)[hv] = pos; + + SET_mmm + + maxLen = 4; + + for (;;) + { + if (d2 < mmm && *(cur - d2) == *cur) + { + distances[0] = 2; + distances[1] = d2 - 1; + distances += 2; + if (*(cur - d2 + 2) == cur[2]) + { + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + distances[1] = d3 - 1; + distances += 2; + d2 = d3; + } + else + break; + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + distances[1] = d3 - 1; + distances += 2; + d2 = d3; + } + else + break; + + distances[-2] = 3; + if (*(cur - d2 + 3) != cur[3]) + break; + UPDATE_maxLen + distances[-2] = maxLen; + if (maxLen == lenLimit) + { + p->son[p->cyclicBufferPos] = curMatch; + MOVE_POS_RET; + } + break; + } + + GET_MATCHES_FOOTER_HC(maxLen); +} + + +UInt32* Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + GET_MATCHES_HEADER(3) + HASH_ZIP_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + GET_MATCHES_FOOTER_HC(2) +} + + +static void Bt2_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + SKIP_HEADER(2) + { + HASH2_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + } + SKIP_FOOTER +} + +void Bt3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + SKIP_HEADER(3) + { + HASH_ZIP_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + } + SKIP_FOOTER +} + +static void Bt3_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + SKIP_HEADER(3) + { + UInt32 h2; + UInt32 *hash; + HASH3_CALC; + hash = p->hash; + curMatch = (hash + kFix3HashSize)[hv]; + hash[h2] = + (hash + kFix3HashSize)[hv] = p->pos; + } + SKIP_FOOTER +} + +static void Bt4_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + SKIP_HEADER(4) + { + UInt32 h2, h3; + UInt32 *hash; + HASH4_CALC; + hash = p->hash; + curMatch = (hash + kFix4HashSize)[hv]; + hash [h2] = + (hash + kFix3HashSize)[h3] = + (hash + kFix4HashSize)[hv] = p->pos; + } + SKIP_FOOTER +} + +static void Bt5_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + SKIP_HEADER(5) + { + UInt32 h2, h3; + UInt32 *hash; + HASH5_CALC; + hash = p->hash; + curMatch = (hash + kFix5HashSize)[hv]; + hash [h2] = + (hash + kFix3HashSize)[h3] = + // (hash + kFix4HashSize)[h4] = + (hash + kFix5HashSize)[hv] = p->pos; + } + SKIP_FOOTER +} + + +#define HC_SKIP_HEADER(minLen) \ + do { if (p->lenLimit < minLen) { MatchFinder_MovePos(p); num--; continue; } { \ + Byte *cur; \ + UInt32 *hash; \ + UInt32 *son; \ + UInt32 pos = p->pos; \ + UInt32 num2 = num; \ + /* (p->pos == p->posLimit) is not allowed here !!! */ \ + { const UInt32 rem = p->posLimit - pos; if (num2 > rem) num2 = rem; } \ + num -= num2; \ + { const UInt32 cycPos = p->cyclicBufferPos; \ + son = p->son + cycPos; \ + p->cyclicBufferPos = cycPos + num2; } \ + cur = p->buffer; \ + hash = p->hash; \ + do { \ + UInt32 curMatch; \ + UInt32 hv; + + +#define HC_SKIP_FOOTER \ + cur++; pos++; *son++ = curMatch; \ + } while (--num2); \ + p->buffer = cur; \ + p->pos = pos; \ + if (pos == p->posLimit) MatchFinder_CheckLimits(p); \ + }} while(num); \ + + +static void Hc4_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + HC_SKIP_HEADER(4) + + UInt32 h2, h3; + HASH4_CALC; + curMatch = (hash + kFix4HashSize)[hv]; + hash [h2] = + (hash + kFix3HashSize)[h3] = + (hash + kFix4HashSize)[hv] = pos; + + HC_SKIP_FOOTER +} + + +static void Hc5_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + HC_SKIP_HEADER(5) + + UInt32 h2, h3; + HASH5_CALC + curMatch = (hash + kFix5HashSize)[hv]; + hash [h2] = + (hash + kFix3HashSize)[h3] = + // (hash + kFix4HashSize)[h4] = + (hash + kFix5HashSize)[hv] = pos; + + HC_SKIP_FOOTER +} + + +void Hc3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + HC_SKIP_HEADER(3) + + HASH_ZIP_CALC; + curMatch = hash[hv]; + hash[hv] = pos; + + HC_SKIP_FOOTER +} + + +void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder2 *vTable) +{ + vTable->Init = (Mf_Init_Func)MatchFinder_Init; + vTable->GetNumAvailableBytes = (Mf_GetNumAvailableBytes_Func)MatchFinder_GetNumAvailableBytes; + vTable->GetPointerToCurrentPos = (Mf_GetPointerToCurrentPos_Func)MatchFinder_GetPointerToCurrentPos; + if (!p->btMode) + { + if (p->numHashBytes <= 4) + { + vTable->GetMatches = (Mf_GetMatches_Func)Hc4_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Hc4_MatchFinder_Skip; + } + else + { + vTable->GetMatches = (Mf_GetMatches_Func)Hc5_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Hc5_MatchFinder_Skip; + } + } + else if (p->numHashBytes == 2) + { + vTable->GetMatches = (Mf_GetMatches_Func)Bt2_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Bt2_MatchFinder_Skip; + } + else if (p->numHashBytes == 3) + { + vTable->GetMatches = (Mf_GetMatches_Func)Bt3_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Bt3_MatchFinder_Skip; + } + else if (p->numHashBytes == 4) + { + vTable->GetMatches = (Mf_GetMatches_Func)Bt4_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Bt4_MatchFinder_Skip; + } + else + { + vTable->GetMatches = (Mf_GetMatches_Func)Bt5_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Bt5_MatchFinder_Skip; + } +} + + + +void LzFindPrepare() +{ + #ifndef FORCE_SATUR_SUB_128 + #ifdef USE_SATUR_SUB_128 + LZFIND_SATUR_SUB_CODE_FUNC f = NULL; + #ifdef MY_CPU_ARM_OR_ARM64 + { + if (CPU_IsSupported_NEON()) + { + // #pragma message ("=== LzFind NEON") + _PRF(printf("\n=== LzFind NEON\n")); + f = LzFind_SaturSub_128; + } + // f = 0; // for debug + } + #else // MY_CPU_ARM_OR_ARM64 + if (CPU_IsSupported_SSE41()) + { + // #pragma message ("=== LzFind SSE41") + _PRF(printf("\n=== LzFind SSE41\n")); + f = LzFind_SaturSub_128; + + #ifdef USE_AVX2 + if (CPU_IsSupported_AVX2()) + { + // #pragma message ("=== LzFind AVX2") + _PRF(printf("\n=== LzFind AVX2\n")); + f = LzFind_SaturSub_256; + } + #endif + } + #endif // MY_CPU_ARM_OR_ARM64 + g_LzFind_SaturSub = f; + #endif // USE_SATUR_SUB_128 + #endif // FORCE_SATUR_SUB_128 +} diff --git a/deps/LZMA-SDK/C/LzFind.h b/deps/LZMA-SDK/C/LzFind.h index 8f9fade23..eea873ff6 100644 --- a/deps/LZMA-SDK/C/LzFind.h +++ b/deps/LZMA-SDK/C/LzFind.h @@ -1,136 +1,136 @@ -/* LzFind.h -- Match finder for LZ algorithms -2021-07-13 : Igor Pavlov : Public domain */ - -#ifndef __LZ_FIND_H -#define __LZ_FIND_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -typedef UInt32 CLzRef; - -typedef struct _CMatchFinder -{ - Byte *buffer; - UInt32 pos; - UInt32 posLimit; - UInt32 streamPos; /* wrap over Zero is allowed (streamPos < pos). Use (UInt32)(streamPos - pos) */ - UInt32 lenLimit; - - UInt32 cyclicBufferPos; - UInt32 cyclicBufferSize; /* it must be = (historySize + 1) */ - - Byte streamEndWasReached; - Byte btMode; - Byte bigHash; - Byte directInput; - - UInt32 matchMaxLen; - CLzRef *hash; - CLzRef *son; - UInt32 hashMask; - UInt32 cutValue; - - Byte *bufferBase; - ISeqInStream *stream; - - UInt32 blockSize; - UInt32 keepSizeBefore; - UInt32 keepSizeAfter; - - UInt32 numHashBytes; - size_t directInputRem; - UInt32 historySize; - UInt32 fixedHashSize; - UInt32 hashSizeSum; - SRes result; - UInt32 crc[256]; - size_t numRefs; - - UInt64 expectedDataSize; -} CMatchFinder; - -#define Inline_MatchFinder_GetPointerToCurrentPos(p) ((const Byte *)(p)->buffer) - -#define Inline_MatchFinder_GetNumAvailableBytes(p) ((UInt32)((p)->streamPos - (p)->pos)) - -/* -#define Inline_MatchFinder_IsFinishedOK(p) \ - ((p)->streamEndWasReached \ - && (p)->streamPos == (p)->pos \ - && (!(p)->directInput || (p)->directInputRem == 0)) -*/ - -int MatchFinder_NeedMove(CMatchFinder *p); -/* Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p); */ -void MatchFinder_MoveBlock(CMatchFinder *p); -void MatchFinder_ReadIfRequired(CMatchFinder *p); - -void MatchFinder_Construct(CMatchFinder *p); - -/* Conditions: - historySize <= 3 GB - keepAddBufferBefore + matchMaxLen + keepAddBufferAfter < 511MB -*/ -int MatchFinder_Create(CMatchFinder *p, UInt32 historySize, - UInt32 keepAddBufferBefore, UInt32 matchMaxLen, UInt32 keepAddBufferAfter, - ISzAllocPtr alloc); -void MatchFinder_Free(CMatchFinder *p, ISzAllocPtr alloc); -void MatchFinder_Normalize3(UInt32 subValue, CLzRef *items, size_t numItems); -// void MatchFinder_ReduceOffsets(CMatchFinder *p, UInt32 subValue); - -/* -#define Inline_MatchFinder_InitPos(p, val) \ - (p)->pos = (val); \ - (p)->streamPos = (val); -*/ - -#define Inline_MatchFinder_ReduceOffsets(p, subValue) \ - (p)->pos -= (subValue); \ - (p)->streamPos -= (subValue); - - -UInt32 * GetMatchesSpec1(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *buffer, CLzRef *son, - size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 _cutValue, - UInt32 *distances, UInt32 maxLen); - -/* -Conditions: - Mf_GetNumAvailableBytes_Func must be called before each Mf_GetMatchLen_Func. - Mf_GetPointerToCurrentPos_Func's result must be used only before any other function -*/ - -typedef void (*Mf_Init_Func)(void *object); -typedef UInt32 (*Mf_GetNumAvailableBytes_Func)(void *object); -typedef const Byte * (*Mf_GetPointerToCurrentPos_Func)(void *object); -typedef UInt32 * (*Mf_GetMatches_Func)(void *object, UInt32 *distances); -typedef void (*Mf_Skip_Func)(void *object, UInt32); - -typedef struct _IMatchFinder -{ - Mf_Init_Func Init; - Mf_GetNumAvailableBytes_Func GetNumAvailableBytes; - Mf_GetPointerToCurrentPos_Func GetPointerToCurrentPos; - Mf_GetMatches_Func GetMatches; - Mf_Skip_Func Skip; -} IMatchFinder2; - -void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder2 *vTable); - -void MatchFinder_Init_LowHash(CMatchFinder *p); -void MatchFinder_Init_HighHash(CMatchFinder *p); -void MatchFinder_Init_4(CMatchFinder *p); -void MatchFinder_Init(CMatchFinder *p); - -UInt32* Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances); -UInt32* Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances); - -void Bt3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num); -void Hc3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num); - -void LzFindPrepare(void); - -EXTERN_C_END - -#endif +/* LzFind.h -- Match finder for LZ algorithms +2021-07-13 : Igor Pavlov : Public domain */ + +#ifndef __LZ_FIND_H +#define __LZ_FIND_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +typedef UInt32 CLzRef; + +typedef struct _CMatchFinder +{ + Byte *buffer; + UInt32 pos; + UInt32 posLimit; + UInt32 streamPos; /* wrap over Zero is allowed (streamPos < pos). Use (UInt32)(streamPos - pos) */ + UInt32 lenLimit; + + UInt32 cyclicBufferPos; + UInt32 cyclicBufferSize; /* it must be = (historySize + 1) */ + + Byte streamEndWasReached; + Byte btMode; + Byte bigHash; + Byte directInput; + + UInt32 matchMaxLen; + CLzRef *hash; + CLzRef *son; + UInt32 hashMask; + UInt32 cutValue; + + Byte *bufferBase; + ISeqInStream *stream; + + UInt32 blockSize; + UInt32 keepSizeBefore; + UInt32 keepSizeAfter; + + UInt32 numHashBytes; + size_t directInputRem; + UInt32 historySize; + UInt32 fixedHashSize; + UInt32 hashSizeSum; + SRes result; + UInt32 crc[256]; + size_t numRefs; + + UInt64 expectedDataSize; +} CMatchFinder; + +#define Inline_MatchFinder_GetPointerToCurrentPos(p) ((const Byte *)(p)->buffer) + +#define Inline_MatchFinder_GetNumAvailableBytes(p) ((UInt32)((p)->streamPos - (p)->pos)) + +/* +#define Inline_MatchFinder_IsFinishedOK(p) \ + ((p)->streamEndWasReached \ + && (p)->streamPos == (p)->pos \ + && (!(p)->directInput || (p)->directInputRem == 0)) +*/ + +int MatchFinder_NeedMove(CMatchFinder *p); +/* Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p); */ +void MatchFinder_MoveBlock(CMatchFinder *p); +void MatchFinder_ReadIfRequired(CMatchFinder *p); + +void MatchFinder_Construct(CMatchFinder *p); + +/* Conditions: + historySize <= 3 GB + keepAddBufferBefore + matchMaxLen + keepAddBufferAfter < 511MB +*/ +int MatchFinder_Create(CMatchFinder *p, UInt32 historySize, + UInt32 keepAddBufferBefore, UInt32 matchMaxLen, UInt32 keepAddBufferAfter, + ISzAllocPtr alloc); +void MatchFinder_Free(CMatchFinder *p, ISzAllocPtr alloc); +void MatchFinder_Normalize3(UInt32 subValue, CLzRef *items, size_t numItems); +// void MatchFinder_ReduceOffsets(CMatchFinder *p, UInt32 subValue); + +/* +#define Inline_MatchFinder_InitPos(p, val) \ + (p)->pos = (val); \ + (p)->streamPos = (val); +*/ + +#define Inline_MatchFinder_ReduceOffsets(p, subValue) \ + (p)->pos -= (subValue); \ + (p)->streamPos -= (subValue); + + +UInt32 * GetMatchesSpec1(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *buffer, CLzRef *son, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 _cutValue, + UInt32 *distances, UInt32 maxLen); + +/* +Conditions: + Mf_GetNumAvailableBytes_Func must be called before each Mf_GetMatchLen_Func. + Mf_GetPointerToCurrentPos_Func's result must be used only before any other function +*/ + +typedef void (*Mf_Init_Func)(void *object); +typedef UInt32 (*Mf_GetNumAvailableBytes_Func)(void *object); +typedef const Byte * (*Mf_GetPointerToCurrentPos_Func)(void *object); +typedef UInt32 * (*Mf_GetMatches_Func)(void *object, UInt32 *distances); +typedef void (*Mf_Skip_Func)(void *object, UInt32); + +typedef struct _IMatchFinder +{ + Mf_Init_Func Init; + Mf_GetNumAvailableBytes_Func GetNumAvailableBytes; + Mf_GetPointerToCurrentPos_Func GetPointerToCurrentPos; + Mf_GetMatches_Func GetMatches; + Mf_Skip_Func Skip; +} IMatchFinder2; + +void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder2 *vTable); + +void MatchFinder_Init_LowHash(CMatchFinder *p); +void MatchFinder_Init_HighHash(CMatchFinder *p); +void MatchFinder_Init_4(CMatchFinder *p); +void MatchFinder_Init(CMatchFinder *p); + +UInt32* Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances); +UInt32* Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances); + +void Bt3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num); +void Hc3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num); + +void LzFindPrepare(void); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/LzFindMt.c b/deps/LZMA-SDK/C/LzFindMt.c index 3865263b7..4e67fc3f2 100644 --- a/deps/LZMA-SDK/C/LzFindMt.c +++ b/deps/LZMA-SDK/C/LzFindMt.c @@ -1,1400 +1,1400 @@ -/* LzFindMt.c -- multithreaded Match finder for LZ algorithms -2021-12-21 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -// #include - -#include "CpuArch.h" - -#include "LzHash.h" -#include "LzFindMt.h" - -// #define LOG_ITERS - -// #define LOG_THREAD - -#ifdef LOG_THREAD -#include -#define PRF(x) x -#else -#define PRF(x) -#endif - -#ifdef LOG_ITERS -#include -extern UInt64 g_NumIters_Tree; -extern UInt64 g_NumIters_Loop; -extern UInt64 g_NumIters_Bytes; -#define LOG_ITER(x) x -#else -#define LOG_ITER(x) -#endif - -#define kMtHashBlockSize ((UInt32)1 << 17) -#define kMtHashNumBlocks (1 << 1) - -#define GET_HASH_BLOCK_OFFSET(i) (((i) & (kMtHashNumBlocks - 1)) * kMtHashBlockSize) - -#define kMtBtBlockSize ((UInt32)1 << 16) -#define kMtBtNumBlocks (1 << 4) - -#define GET_BT_BLOCK_OFFSET(i) (((i) & (kMtBtNumBlocks - 1)) * (size_t)kMtBtBlockSize) - -/* - HASH functions: - We use raw 8/16 bits from a[1] and a[2], - xored with crc(a[0]) and crc(a[3]). - We check a[0], a[3] only. We don't need to compare a[1] and a[2] in matches. - our crc() function provides one-to-one correspondence for low 8-bit values: - (crc[0...0xFF] & 0xFF) <-> [0...0xFF] -*/ - -#define MF(mt) ((mt)->MatchFinder) -#define MF_CRC (p->crc) - -// #define MF(mt) (&(mt)->MatchFinder) -// #define MF_CRC (p->MatchFinder.crc) - -#define MT_HASH2_CALC \ - h2 = (MF_CRC[cur[0]] ^ cur[1]) & (kHash2Size - 1); - -#define MT_HASH3_CALC { \ - UInt32 temp = MF_CRC[cur[0]] ^ cur[1]; \ - h2 = temp & (kHash2Size - 1); \ - h3 = (temp ^ ((UInt32)cur[2] << 8)) & (kHash3Size - 1); } - -/* -#define MT_HASH3_CALC__NO_2 { \ - UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ - h3 = (temp ^ ((UInt32)cur[2] << 8)) & (kHash3Size - 1); } - -#define __MT_HASH4_CALC { \ - UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ - h2 = temp & (kHash2Size - 1); \ - temp ^= ((UInt32)cur[2] << 8); \ - h3 = temp & (kHash3Size - 1); \ - h4 = (temp ^ (p->crc[cur[3]] << kLzHash_CrcShift_1)) & p->hash4Mask; } - // (kHash4Size - 1); -*/ - - -MY_NO_INLINE -static void MtSync_Construct(CMtSync *p) -{ - p->affinity = 0; - p->wasCreated = False; - p->csWasInitialized = False; - p->csWasEntered = False; - Thread_Construct(&p->thread); - Event_Construct(&p->canStart); - Event_Construct(&p->wasStopped); - Semaphore_Construct(&p->freeSemaphore); - Semaphore_Construct(&p->filledSemaphore); -} - - -#define DEBUG_BUFFER_LOCK // define it to debug lock state - -#ifdef DEBUG_BUFFER_LOCK -#include -#define BUFFER_MUST_BE_LOCKED(p) if (!(p)->csWasEntered) exit(1); -#define BUFFER_MUST_BE_UNLOCKED(p) if ( (p)->csWasEntered) exit(1); -#else -#define BUFFER_MUST_BE_LOCKED(p) -#define BUFFER_MUST_BE_UNLOCKED(p) -#endif - -#define LOCK_BUFFER(p) { \ - BUFFER_MUST_BE_UNLOCKED(p); \ - CriticalSection_Enter(&(p)->cs); \ - (p)->csWasEntered = True; } - -#define UNLOCK_BUFFER(p) { \ - BUFFER_MUST_BE_LOCKED(p); \ - CriticalSection_Leave(&(p)->cs); \ - (p)->csWasEntered = False; } - - -MY_NO_INLINE -static UInt32 MtSync_GetNextBlock(CMtSync *p) -{ - UInt32 numBlocks = 0; - if (p->needStart) - { - BUFFER_MUST_BE_UNLOCKED(p) - p->numProcessedBlocks = 1; - p->needStart = False; - p->stopWriting = False; - p->exit = False; - Event_Reset(&p->wasStopped); - Event_Set(&p->canStart); - } - else - { - UNLOCK_BUFFER(p) - // we free current block - numBlocks = p->numProcessedBlocks++; - Semaphore_Release1(&p->freeSemaphore); - } - - // buffer is UNLOCKED here - Semaphore_Wait(&p->filledSemaphore); - LOCK_BUFFER(p); - return numBlocks; -} - - -/* if Writing (Processing) thread was started, we must call MtSync_StopWriting() */ - -MY_NO_INLINE -static void MtSync_StopWriting(CMtSync *p) -{ - if (!Thread_WasCreated(&p->thread) || p->needStart) - return; - - PRF(printf("\nMtSync_StopWriting %p\n", p)); - - if (p->csWasEntered) - { - /* we don't use buffer in this thread after StopWriting(). - So we UNLOCK buffer. - And we restore default UNLOCKED state for stopped thread */ - UNLOCK_BUFFER(p) - } - - /* We send (p->stopWriting) message and release freeSemaphore - to free current block. - So the thread will see (p->stopWriting) at some - iteration after Wait(freeSemaphore). - The thread doesn't need to fill all avail free blocks, - so we can get fast thread stop. - */ - - p->stopWriting = True; - Semaphore_Release1(&p->freeSemaphore); // check semaphore count !!! - - PRF(printf("\nMtSync_StopWriting %p : Event_Wait(&p->wasStopped)\n", p)); - Event_Wait(&p->wasStopped); - PRF(printf("\nMtSync_StopWriting %p : Event_Wait() finsihed\n", p)); - - /* 21.03 : we don't restore samaphore counters here. - We will recreate and reinit samaphores in next start */ - - p->needStart = True; -} - - -MY_NO_INLINE -static void MtSync_Destruct(CMtSync *p) -{ - PRF(printf("\nMtSync_Destruct %p\n", p)); - - if (Thread_WasCreated(&p->thread)) - { - /* we want thread to be in Stopped state before sending EXIT command. - note: stop(btSync) will stop (htSync) also */ - MtSync_StopWriting(p); - /* thread in Stopped state here : (p->needStart == true) */ - p->exit = True; - // if (p->needStart) // it's (true) - Event_Set(&p->canStart); // we send EXIT command to thread - Thread_Wait_Close(&p->thread); // we wait thread finishing - } - - if (p->csWasInitialized) - { - CriticalSection_Delete(&p->cs); - p->csWasInitialized = False; - } - p->csWasEntered = False; - - Event_Close(&p->canStart); - 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; } -// we want to get real system error codes here instead of SZ_ERROR_THREAD -#define RINOK_THREAD(x) RINOK(x) - - -// call it before each new file (when new starting is required): -MY_NO_INLINE -static SRes MtSync_Init(CMtSync *p, UInt32 numBlocks) -{ - WRes wres; - // BUFFER_MUST_BE_UNLOCKED(p) - if (!p->needStart || p->csWasEntered) - return SZ_ERROR_FAIL; - wres = Semaphore_OptCreateInit(&p->freeSemaphore, numBlocks, numBlocks); - if (wres == 0) - wres = Semaphore_OptCreateInit(&p->filledSemaphore, 0, numBlocks); - return MY_SRes_HRESULT_FROM_WRes(wres); -} - - -static WRes MtSync_Create_WRes(CMtSync *p, THREAD_FUNC_TYPE startAddress, void *obj) -{ - WRes wres; - - if (p->wasCreated) - return SZ_OK; - - RINOK_THREAD(CriticalSection_Init(&p->cs)); - p->csWasInitialized = True; - p->csWasEntered = False; - - RINOK_THREAD(AutoResetEvent_CreateNotSignaled(&p->canStart)); - RINOK_THREAD(AutoResetEvent_CreateNotSignaled(&p->wasStopped)); - - p->needStart = True; - p->exit = True; /* p->exit is unused before (canStart) Event. - But in case of some unexpected code failure we will get fast exit from thread */ - - // return ERROR_TOO_MANY_POSTS; // for debug - // return EINVAL; // for debug - - if (p->affinity != 0) - wres = Thread_Create_With_Affinity(&p->thread, startAddress, obj, (CAffinityMask)p->affinity); - else - wres = Thread_Create(&p->thread, startAddress, obj); - - RINOK_THREAD(wres); - p->wasCreated = True; - return SZ_OK; -} - - -MY_NO_INLINE -static SRes MtSync_Create(CMtSync *p, THREAD_FUNC_TYPE startAddress, void *obj) -{ - const WRes wres = MtSync_Create_WRes(p, startAddress, obj); - if (wres == 0) - return 0; - MtSync_Destruct(p); - return MY_SRes_HRESULT_FROM_WRes(wres); -} - - -// ---------- HASH THREAD ---------- - -#define kMtMaxValForNormalize 0xFFFFFFFF -// #define kMtMaxValForNormalize ((1 << 21)) // for debug -// #define kNormalizeAlign (1 << 7) // alignment for speculated accesses - -#ifdef MY_CPU_LE_UNALIGN - #define GetUi24hi_from32(p) ((UInt32)GetUi32(p) >> 8) -#else - #define GetUi24hi_from32(p) ((p)[1] ^ ((UInt32)(p)[2] << 8) ^ ((UInt32)(p)[3] << 16)) -#endif - -#define GetHeads_DECL(name) \ - static void GetHeads ## name(const Byte *p, UInt32 pos, \ - UInt32 *hash, UInt32 hashMask, UInt32 *heads, UInt32 numHeads, const UInt32 *crc) - -#define GetHeads_LOOP(v) \ - for (; numHeads != 0; numHeads--) { \ - const UInt32 value = (v); \ - p++; \ - *heads++ = pos - hash[value]; \ - hash[value] = pos++; } - -#define DEF_GetHeads2(name, v, action) \ - GetHeads_DECL(name) { action \ - GetHeads_LOOP(v) } - -#define DEF_GetHeads(name, v) DEF_GetHeads2(name, v, ;) - -DEF_GetHeads2(2, GetUi16(p), UNUSED_VAR(hashMask); UNUSED_VAR(crc); ) -DEF_GetHeads(3, (crc[p[0]] ^ GetUi16(p + 1)) & hashMask) -DEF_GetHeads2(3b, GetUi16(p) ^ ((UInt32)(p)[2] << 16), UNUSED_VAR(hashMask); UNUSED_VAR(crc); ) -// BT3 is not good for crc collisions for big hashMask values. - -/* -GetHeads_DECL(3b) -{ - UNUSED_VAR(hashMask); - UNUSED_VAR(crc); - { - const Byte *pLim = p + numHeads; - if (numHeads == 0) - return; - pLim--; - while (p < pLim) - { - UInt32 v1 = GetUi32(p); - UInt32 v0 = v1 & 0xFFFFFF; - UInt32 h0, h1; - p += 2; - v1 >>= 8; - h0 = hash[v0]; hash[v0] = pos; heads[0] = pos - h0; pos++; - h1 = hash[v1]; hash[v1] = pos; heads[1] = pos - h1; pos++; - heads += 2; - } - if (p == pLim) - { - UInt32 v0 = GetUi16(p) ^ ((UInt32)(p)[2] << 16); - *heads = pos - hash[v0]; - hash[v0] = pos; - } - } -} -*/ - -/* -GetHeads_DECL(4) -{ - unsigned sh = 0; - UNUSED_VAR(crc) - while ((hashMask & 0x80000000) == 0) - { - hashMask <<= 1; - sh++; - } - GetHeads_LOOP((GetUi32(p) * 0xa54a1) >> sh) -} -#define GetHeads4b GetHeads4 -*/ - -#define USE_GetHeads_LOCAL_CRC - -#ifdef USE_GetHeads_LOCAL_CRC - -GetHeads_DECL(4) -{ - UInt32 crc0[256]; - UInt32 crc1[256]; - { - unsigned i; - for (i = 0; i < 256; i++) - { - UInt32 v = crc[i]; - crc0[i] = v & hashMask; - crc1[i] = (v << kLzHash_CrcShift_1) & hashMask; - // crc1[i] = rotlFixed(v, 8) & hashMask; - } - } - GetHeads_LOOP(crc0[p[0]] ^ crc1[p[3]] ^ (UInt32)GetUi16(p+1)) -} - -GetHeads_DECL(4b) -{ - UInt32 crc0[256]; - { - unsigned i; - for (i = 0; i < 256; i++) - crc0[i] = crc[i] & hashMask; - } - GetHeads_LOOP(crc0[p[0]] ^ GetUi24hi_from32(p)) -} - -GetHeads_DECL(5) -{ - UInt32 crc0[256]; - UInt32 crc1[256]; - UInt32 crc2[256]; - { - unsigned i; - for (i = 0; i < 256; i++) - { - UInt32 v = crc[i]; - crc0[i] = v & hashMask; - crc1[i] = (v << kLzHash_CrcShift_1) & hashMask; - crc2[i] = (v << kLzHash_CrcShift_2) & hashMask; - } - } - GetHeads_LOOP(crc0[p[0]] ^ crc1[p[3]] ^ crc2[p[4]] ^ (UInt32)GetUi16(p+1)) -} - -GetHeads_DECL(5b) -{ - UInt32 crc0[256]; - UInt32 crc1[256]; - { - unsigned i; - for (i = 0; i < 256; i++) - { - UInt32 v = crc[i]; - crc0[i] = v & hashMask; - crc1[i] = (v << kLzHash_CrcShift_1) & hashMask; - } - } - GetHeads_LOOP(crc0[p[0]] ^ crc1[p[4]] ^ GetUi24hi_from32(p)) -} - -#else - -DEF_GetHeads(4, (crc[p[0]] ^ (crc[p[3]] << kLzHash_CrcShift_1) ^ (UInt32)GetUi16(p+1)) & hashMask) -DEF_GetHeads(4b, (crc[p[0]] ^ GetUi24hi_from32(p)) & hashMask) -DEF_GetHeads(5, (crc[p[0]] ^ (crc[p[3]] << kLzHash_CrcShift_1) ^ (crc[p[4]] << kLzHash_CrcShift_2) ^ (UInt32)GetUi16(p + 1)) & hashMask) -DEF_GetHeads(5b, (crc[p[0]] ^ (crc[p[4]] << kLzHash_CrcShift_1) ^ GetUi24hi_from32(p)) & hashMask) - -#endif - - -static void HashThreadFunc(CMatchFinderMt *mt) -{ - CMtSync *p = &mt->hashSync; - PRF(printf("\nHashThreadFunc\n")); - - for (;;) - { - UInt32 blockIndex = 0; - PRF(printf("\nHashThreadFunc : Event_Wait(&p->canStart)\n")); - Event_Wait(&p->canStart); - PRF(printf("\nHashThreadFunc : Event_Wait(&p->canStart) : after \n")); - if (p->exit) - { - PRF(printf("\nHashThreadFunc : exit \n")); - return; - } - - MatchFinder_Init_HighHash(MF(mt)); - - for (;;) - { - PRF(printf("Hash thread block = %d pos = %d\n", (unsigned)blockIndex, mt->MatchFinder->pos)); - - { - CMatchFinder *mf = MF(mt); - if (MatchFinder_NeedMove(mf)) - { - CriticalSection_Enter(&mt->btSync.cs); - CriticalSection_Enter(&mt->hashSync.cs); - { - const Byte *beforePtr = Inline_MatchFinder_GetPointerToCurrentPos(mf); - ptrdiff_t offset; - MatchFinder_MoveBlock(mf); - offset = beforePtr - Inline_MatchFinder_GetPointerToCurrentPos(mf); - mt->pointerToCurPos -= offset; - mt->buffer -= offset; - } - CriticalSection_Leave(&mt->hashSync.cs); - CriticalSection_Leave(&mt->btSync.cs); - continue; - } - - Semaphore_Wait(&p->freeSemaphore); - - if (p->exit) // exit is unexpected here. But we check it here for some failure case - return; - - // for faster stop : we check (p->stopWriting) after Wait(freeSemaphore) - if (p->stopWriting) - break; - - MatchFinder_ReadIfRequired(mf); - { - UInt32 *heads = mt->hashBuf + GET_HASH_BLOCK_OFFSET(blockIndex++); - UInt32 num = Inline_MatchFinder_GetNumAvailableBytes(mf); - heads[0] = 2; - heads[1] = num; - - /* heads[1] contains the number of avail bytes: - if (avail < mf->numHashBytes) : - { - it means that stream was finished - HASH_THREAD and BT_TREAD must move position for heads[1] (avail) bytes. - HASH_THREAD doesn't stop, - HASH_THREAD fills only the header (2 numbers) for all next blocks: - {2, NumHashBytes - 1}, {2,0}, {2,0}, ... , {2,0} - } - else - { - HASH_THREAD and BT_TREAD must move position for (heads[0] - 2) bytes; - } - */ - - if (num >= mf->numHashBytes) - { - num = num - mf->numHashBytes + 1; - if (num > kMtHashBlockSize - 2) - num = kMtHashBlockSize - 2; - - if (mf->pos > (UInt32)kMtMaxValForNormalize - num) - { - const UInt32 subValue = (mf->pos - mf->historySize - 1); // & ~(UInt32)(kNormalizeAlign - 1); - Inline_MatchFinder_ReduceOffsets(mf, subValue); - MatchFinder_Normalize3(subValue, mf->hash + mf->fixedHashSize, (size_t)mf->hashMask + 1); - } - - heads[0] = 2 + num; - mt->GetHeadsFunc(mf->buffer, mf->pos, mf->hash + mf->fixedHashSize, mf->hashMask, heads + 2, num, mf->crc); - } - - mf->pos += num; // wrap over zero is allowed at the end of stream - mf->buffer += num; - } - } - - Semaphore_Release1(&p->filledSemaphore); - } // for() processing end - - // p->numBlocks_Sent = blockIndex; - Event_Set(&p->wasStopped); - } // for() thread end -} - - - - -// ---------- BT THREAD ---------- - -/* we use one variable instead of two (cyclicBufferPos == pos) before CyclicBuf wrap. - here we define fixed offset of (p->pos) from (p->cyclicBufferPos) */ -#define CYC_TO_POS_OFFSET 0 -// #define CYC_TO_POS_OFFSET 1 // for debug - -#define MFMT_GM_INLINE - -#ifdef MFMT_GM_INLINE - -/* - we use size_t for (pos) instead of UInt32 - to eliminate "movsx" BUG in old MSVC x64 compiler. -*/ - - -UInt32 * MY_FAST_CALL GetMatchesSpecN_2(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, - UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, - size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, - UInt32 *posRes); - -#endif - - -static void BtGetMatches(CMatchFinderMt *p, UInt32 *d) -{ - UInt32 numProcessed = 0; - UInt32 curPos = 2; - - /* GetMatchesSpec() functions don't create (len = 1) - in [len, dist] match pairs, if (p->numHashBytes >= 2) - Also we suppose here that (matchMaxLen >= 2). - So the following code for (reserve) is not required - UInt32 reserve = (p->matchMaxLen * 2); - const UInt32 kNumHashBytes_Max = 5; // BT_HASH_BYTES_MAX - if (reserve < kNumHashBytes_Max - 1) - reserve = kNumHashBytes_Max - 1; - const UInt32 limit = kMtBtBlockSize - (reserve); - */ - - const UInt32 limit = kMtBtBlockSize - (p->matchMaxLen * 2); - - d[1] = p->hashNumAvail; - - if (p->failure_BT) - { - // printf("\n == 1 BtGetMatches() p->failure_BT\n"); - d[0] = 0; - // d[1] = 0; - return; - } - - while (curPos < limit) - { - if (p->hashBufPos == p->hashBufPosLimit) - { - // MatchFinderMt_GetNextBlock_Hash(p); - UInt32 avail; - { - const UInt32 bi = MtSync_GetNextBlock(&p->hashSync); - const UInt32 k = GET_HASH_BLOCK_OFFSET(bi); - const UInt32 *h = p->hashBuf + k; - avail = h[1]; - p->hashBufPosLimit = k + h[0]; - p->hashNumAvail = avail; - p->hashBufPos = k + 2; - } - - { - /* we must prevent UInt32 overflow for avail total value, - if avail was increased with new hash block */ - UInt32 availSum = numProcessed + avail; - if (availSum < numProcessed) - availSum = (UInt32)(Int32)-1; - d[1] = availSum; - } - - if (avail >= p->numHashBytes) - continue; - - // if (p->hashBufPos != p->hashBufPosLimit) exit(1); - - /* (avail < p->numHashBytes) - It means that stream was finished. - And (avail) - is a number of remaining bytes, - we fill (d) for (avail) bytes for LZ_THREAD (receiver). - but we don't update (p->pos) and (p->cyclicBufferPos) here in BT_THREAD */ - - /* here we suppose that we have space enough: - (kMtBtBlockSize - curPos >= p->hashNumAvail) */ - p->hashNumAvail = 0; - d[0] = curPos + avail; - d += curPos; - for (; avail != 0; avail--) - *d++ = 0; - return; - } - { - UInt32 size = p->hashBufPosLimit - p->hashBufPos; - UInt32 pos = p->pos; - UInt32 cyclicBufferPos = p->cyclicBufferPos; - UInt32 lenLimit = p->matchMaxLen; - 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; - } - - if (pos > (UInt32)kMtMaxValForNormalize - size) - { - const UInt32 subValue = (pos - p->cyclicBufferSize); // & ~(UInt32)(kNormalizeAlign - 1); - pos -= subValue; - p->pos = pos; - MatchFinder_Normalize3(subValue, p->son, (size_t)p->cyclicBufferSize * 2); - } - - #ifndef MFMT_GM_INLINE - while (curPos < limit && size-- != 0) - { - UInt32 *startDistances = d + 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 = pos; - const UInt32 *d_end; - { - d_end = GetMatchesSpecN_2( - p->buffer + lenLimit - 1, - pos, p->buffer, p->son, p->cutValue, d + curPos, - p->numHashBytes - 1, p->hashBuf + p->hashBufPos, - d + limit, p->hashBuf + p->hashBufPos + size, - cyclicBufferPos, p->cyclicBufferSize, - &posRes); - } - { - if (!d_end) - { - // printf("\n == 2 BtGetMatches() p->failure_BT\n"); - // internal data failure - p->failure_BT = True; - d[0] = 0; - // d[1] = 0; - return; - } - } - curPos = (UInt32)(d_end - d); - { - const UInt32 processed = posRes - pos; - pos = posRes; - p->hashBufPos += processed; - cyclicBufferPos += processed; - p->buffer += processed; - } - } - #endif - - { - const UInt32 processed = pos - p->pos; - numProcessed += processed; - p->hashNumAvail -= processed; - p->pos = pos; - } - if (cyclicBufferPos == p->cyclicBufferSize) - cyclicBufferPos = 0; - p->cyclicBufferPos = cyclicBufferPos; - } - } - - d[0] = curPos; -} - - -static void BtFillBlock(CMatchFinderMt *p, UInt32 globalBlockIndex) -{ - CMtSync *sync = &p->hashSync; - - BUFFER_MUST_BE_UNLOCKED(sync) - - if (!sync->needStart) - { - LOCK_BUFFER(sync) - } - - BtGetMatches(p, p->btBuf + GET_BT_BLOCK_OFFSET(globalBlockIndex)); - - /* We suppose that we have called GetNextBlock() from start. - So buffer is LOCKED */ - - UNLOCK_BUFFER(sync) -} - - -MY_NO_INLINE -static void BtThreadFunc(CMatchFinderMt *mt) -{ - CMtSync *p = &mt->btSync; - for (;;) - { - UInt32 blockIndex = 0; - Event_Wait(&p->canStart); - - for (;;) - { - PRF(printf(" BT thread block = %d pos = %d\n", (unsigned)blockIndex, mt->pos)); - /* (p->exit == true) is possible after (p->canStart) at first loop iteration - and is unexpected after more Wait(freeSemaphore) iterations */ - if (p->exit) - return; - - Semaphore_Wait(&p->freeSemaphore); - - // for faster stop : we check (p->stopWriting) after Wait(freeSemaphore) - if (p->stopWriting) - break; - - BtFillBlock(mt, blockIndex++); - - Semaphore_Release1(&p->filledSemaphore); - } - - // we stop HASH_THREAD here - MtSync_StopWriting(&mt->hashSync); - - // p->numBlocks_Sent = blockIndex; - Event_Set(&p->wasStopped); - } -} - - -void MatchFinderMt_Construct(CMatchFinderMt *p) -{ - p->hashBuf = NULL; - MtSync_Construct(&p->hashSync); - MtSync_Construct(&p->btSync); -} - -static void MatchFinderMt_FreeMem(CMatchFinderMt *p, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, p->hashBuf); - p->hashBuf = NULL; -} - -void MatchFinderMt_Destruct(CMatchFinderMt *p, ISzAllocPtr alloc) -{ - /* - HASH_THREAD can use CriticalSection(s) btSync.cs and hashSync.cs. - So we must be sure that HASH_THREAD will not use CriticalSection(s) - after deleting CriticalSection here. - - we call ReleaseStream(p) - that calls StopWriting(btSync) - that calls StopWriting(hashSync), if it's required to stop HASH_THREAD. - after StopWriting() it's safe to destruct MtSync(s) in any order */ - - MatchFinderMt_ReleaseStream(p); - - MtSync_Destruct(&p->btSync); - MtSync_Destruct(&p->hashSync); - - LOG_ITER( - printf("\nTree %9d * %7d iter = %9d = sum : bytes = %9d\n", - (UInt32)(g_NumIters_Tree / 1000), - (UInt32)(((UInt64)g_NumIters_Loop * 1000) / (g_NumIters_Tree + 1)), - (UInt32)(g_NumIters_Loop / 1000), - (UInt32)(g_NumIters_Bytes / 1000) - )); - - MatchFinderMt_FreeMem(p, alloc); -} - - -#define kHashBufferSize (kMtHashBlockSize * kMtHashNumBlocks) -#define kBtBufferSize (kMtBtBlockSize * kMtBtNumBlocks) - - -static THREAD_FUNC_DECL HashThreadFunc2(void *p) { HashThreadFunc((CMatchFinderMt *)p); return 0; } -static THREAD_FUNC_DECL BtThreadFunc2(void *p) -{ - Byte allocaDummy[0x180]; - unsigned i = 0; - for (i = 0; i < 16; i++) - allocaDummy[i] = (Byte)0; - if (allocaDummy[0] == 0) - BtThreadFunc((CMatchFinderMt *)p); - return 0; -} - - -SRes MatchFinderMt_Create(CMatchFinderMt *p, UInt32 historySize, UInt32 keepAddBufferBefore, - UInt32 matchMaxLen, UInt32 keepAddBufferAfter, ISzAllocPtr alloc) -{ - CMatchFinder *mf = MF(p); - p->historySize = historySize; - if (kMtBtBlockSize <= matchMaxLen * 4) - return SZ_ERROR_PARAM; - if (!p->hashBuf) - { - p->hashBuf = (UInt32 *)ISzAlloc_Alloc(alloc, ((size_t)kHashBufferSize + (size_t)kBtBufferSize) * sizeof(UInt32)); - if (!p->hashBuf) - return SZ_ERROR_MEM; - p->btBuf = p->hashBuf + kHashBufferSize; - } - keepAddBufferBefore += (kHashBufferSize + kBtBufferSize); - keepAddBufferAfter += kMtHashBlockSize; - if (!MatchFinder_Create(mf, historySize, keepAddBufferBefore, matchMaxLen, keepAddBufferAfter, alloc)) - return SZ_ERROR_MEM; - - RINOK(MtSync_Create(&p->hashSync, HashThreadFunc2, p)); - RINOK(MtSync_Create(&p->btSync, BtThreadFunc2, p)); - return SZ_OK; -} - - -SRes MatchFinderMt_InitMt(CMatchFinderMt *p) -{ - RINOK(MtSync_Init(&p->hashSync, kMtHashNumBlocks)); - return MtSync_Init(&p->btSync, kMtBtNumBlocks); -} - - -static void MatchFinderMt_Init(CMatchFinderMt *p) -{ - CMatchFinder *mf = MF(p); - - p->btBufPos = - p->btBufPosLimit = NULL; - p->hashBufPos = - p->hashBufPosLimit = 0; - p->hashNumAvail = 0; // 21.03 - - p->failure_BT = False; - - /* Init without data reading. We don't want to read data in this thread */ - MatchFinder_Init_4(mf); - - MatchFinder_Init_LowHash(mf); - - p->pointerToCurPos = Inline_MatchFinder_GetPointerToCurrentPos(mf); - p->btNumAvailBytes = 0; - p->failure_LZ_BT = False; - // p->failure_LZ_LZ = False; - - p->lzPos = - 1; // optimal smallest value - // 0; // for debug: ignores match to start - // kNormalizeAlign; // for debug - - p->hash = mf->hash; - p->fixedHashSize = mf->fixedHashSize; - // p->hash4Mask = mf->hash4Mask; - p->crc = mf->crc; - // memcpy(p->crc, mf->crc, sizeof(mf->crc)); - - p->son = mf->son; - p->matchMaxLen = mf->matchMaxLen; - p->numHashBytes = mf->numHashBytes; - - /* (mf->pos) and (mf->streamPos) were already initialized to 1 in MatchFinder_Init_4() */ - // mf->streamPos = mf->pos = 1; // optimal smallest value - // 0; // for debug: ignores match to start - // kNormalizeAlign; // for debug - - /* we must init (p->pos = mf->pos) for BT, because - BT code needs (p->pos == delta_value_for_empty_hash_record == mf->pos) */ - p->pos = mf->pos; // do not change it - - p->cyclicBufferPos = (p->pos - CYC_TO_POS_OFFSET); - p->cyclicBufferSize = mf->cyclicBufferSize; - p->buffer = mf->buffer; - p->cutValue = mf->cutValue; - // p->son[0] = p->son[1] = 0; // unused: to init skipped record for speculated accesses. -} - - -/* ReleaseStream is required to finish multithreading */ -void MatchFinderMt_ReleaseStream(CMatchFinderMt *p) -{ - // Sleep(1); // for debug - MtSync_StopWriting(&p->btSync); - // Sleep(200); // for debug - /* p->MatchFinder->ReleaseStream(); */ -} - - -MY_NO_INLINE -static UInt32 MatchFinderMt_GetNextBlock_Bt(CMatchFinderMt *p) -{ - if (p->failure_LZ_BT) - p->btBufPos = p->failureBuf; - else - { - const UInt32 bi = MtSync_GetNextBlock(&p->btSync); - const UInt32 *bt = p->btBuf + GET_BT_BLOCK_OFFSET(bi); - { - const UInt32 numItems = bt[0]; - p->btBufPosLimit = bt + numItems; - p->btNumAvailBytes = bt[1]; - p->btBufPos = bt + 2; - if (numItems < 2 || numItems > kMtBtBlockSize) - { - p->failureBuf[0] = 0; - p->btBufPos = p->failureBuf; - p->btBufPosLimit = p->failureBuf + 1; - p->failure_LZ_BT = True; - // p->btNumAvailBytes = 0; - /* we don't want to decrease AvailBytes, that was load before. - that can be unxepected for the code that have loaded anopther value before */ - } - } - - if (p->lzPos >= (UInt32)kMtMaxValForNormalize - (UInt32)kMtBtBlockSize) - { - /* we don't check (lzPos) over exact avail bytes in (btBuf). - (fixedHashSize) is small, so normalization is fast */ - const UInt32 subValue = (p->lzPos - p->historySize - 1); // & ~(UInt32)(kNormalizeAlign - 1); - p->lzPos -= subValue; - MatchFinder_Normalize3(subValue, p->hash, p->fixedHashSize); - } - } - return p->btNumAvailBytes; -} - - - -static const Byte * MatchFinderMt_GetPointerToCurrentPos(CMatchFinderMt *p) -{ - return p->pointerToCurPos; -} - - -#define GET_NEXT_BLOCK_IF_REQUIRED if (p->btBufPos == p->btBufPosLimit) MatchFinderMt_GetNextBlock_Bt(p); - - -static UInt32 MatchFinderMt_GetNumAvailableBytes(CMatchFinderMt *p) -{ - if (p->btBufPos != p->btBufPosLimit) - return p->btNumAvailBytes; - return MatchFinderMt_GetNextBlock_Bt(p); -} - - -// #define CHECK_FAILURE_LZ(_match_, _pos_) if (_match_ >= _pos_) { p->failure_LZ_LZ = True; return d; } -#define CHECK_FAILURE_LZ(_match_, _pos_) - -static UInt32 * MixMatches2(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *d) -{ - UInt32 h2, c2; - UInt32 *hash = p->hash; - const Byte *cur = p->pointerToCurPos; - const UInt32 m = p->lzPos; - MT_HASH2_CALC - - c2 = hash[h2]; - hash[h2] = m; - - if (c2 >= matchMinPos) - { - CHECK_FAILURE_LZ(c2, m) - if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m] == cur[0]) - { - *d++ = 2; - *d++ = m - c2 - 1; - } - } - - return d; -} - -static UInt32 * MixMatches3(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *d) -{ - UInt32 h2, h3, c2, c3; - UInt32 *hash = p->hash; - const Byte *cur = p->pointerToCurPos; - const UInt32 m = p->lzPos; - MT_HASH3_CALC - - c2 = hash[h2]; - c3 = (hash + kFix3HashSize)[h3]; - - hash[h2] = m; - (hash + kFix3HashSize)[h3] = m; - - if (c2 >= matchMinPos) - { - CHECK_FAILURE_LZ(c2, m) - if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m] == cur[0]) - { - d[1] = m - c2 - 1; - if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m + 2] == cur[2]) - { - d[0] = 3; - return d + 2; - } - d[0] = 2; - d += 2; - } - } - - if (c3 >= matchMinPos) - { - CHECK_FAILURE_LZ(c3, m) - if (cur[(ptrdiff_t)c3 - (ptrdiff_t)m] == cur[0]) - { - *d++ = 3; - *d++ = m - c3 - 1; - } - } - - return d; -} - - -#define INCREASE_LZ_POS p->lzPos++; p->pointerToCurPos++; - -/* -static -UInt32* MatchFinderMt_GetMatches_Bt4(CMatchFinderMt *p, UInt32 *d) -{ - const UInt32 *bt = p->btBufPos; - const UInt32 len = *bt++; - const UInt32 *btLim = bt + len; - UInt32 matchMinPos; - UInt32 avail = p->btNumAvailBytes - 1; - p->btBufPos = btLim; - - { - p->btNumAvailBytes = avail; - - #define BT_HASH_BYTES_MAX 5 - - matchMinPos = p->lzPos; - - if (len != 0) - matchMinPos -= bt[1]; - else if (avail < (BT_HASH_BYTES_MAX - 1) - 1) - { - INCREASE_LZ_POS - return d; - } - else - { - const UInt32 hs = p->historySize; - if (matchMinPos > hs) - matchMinPos -= hs; - else - matchMinPos = 1; - } - } - - for (;;) - { - - UInt32 h2, h3, c2, c3; - UInt32 *hash = p->hash; - const Byte *cur = p->pointerToCurPos; - UInt32 m = p->lzPos; - MT_HASH3_CALC - - c2 = hash[h2]; - c3 = (hash + kFix3HashSize)[h3]; - - hash[h2] = m; - (hash + kFix3HashSize)[h3] = m; - - if (c2 >= matchMinPos && cur[(ptrdiff_t)c2 - (ptrdiff_t)m] == cur[0]) - { - d[1] = m - c2 - 1; - if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m + 2] == cur[2]) - { - d[0] = 3; - d += 2; - break; - } - // else - { - d[0] = 2; - d += 2; - } - } - if (c3 >= matchMinPos && cur[(ptrdiff_t)c3 - (ptrdiff_t)m] == cur[0]) - { - *d++ = 3; - *d++ = m - c3 - 1; - } - break; - } - - if (len != 0) - { - do - { - const UInt32 v0 = bt[0]; - const UInt32 v1 = bt[1]; - bt += 2; - d[0] = v0; - d[1] = v1; - d += 2; - } - while (bt != btLim); - } - INCREASE_LZ_POS - return d; -} -*/ - - -static UInt32 *MixMatches4(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *d) -{ - UInt32 h2, h3, /* h4, */ c2, c3 /* , c4 */; - UInt32 *hash = p->hash; - const Byte *cur = p->pointerToCurPos; - const UInt32 m = p->lzPos; - MT_HASH3_CALC - // MT_HASH4_CALC - c2 = hash[h2]; - c3 = (hash + kFix3HashSize)[h3]; - // c4 = (hash + kFix4HashSize)[h4]; - - hash[h2] = m; - (hash + kFix3HashSize)[h3] = m; - // (hash + kFix4HashSize)[h4] = m; - - #define _USE_H2 - - #ifdef _USE_H2 - if (c2 >= matchMinPos && cur[(ptrdiff_t)c2 - (ptrdiff_t)m] == cur[0]) - { - d[1] = m - c2 - 1; - if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m + 2] == cur[2]) - { - // d[0] = (cur[(ptrdiff_t)c2 - (ptrdiff_t)m + 3] == cur[3]) ? 4 : 3; - // return d + 2; - - if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m + 3] == cur[3]) - { - d[0] = 4; - return d + 2; - } - d[0] = 3; - d += 2; - - #ifdef _USE_H4 - if (c4 >= matchMinPos) - if ( - cur[(ptrdiff_t)c4 - (ptrdiff_t)m] == cur[0] && - cur[(ptrdiff_t)c4 - (ptrdiff_t)m + 3] == cur[3] - ) - { - *d++ = 4; - *d++ = m - c4 - 1; - } - #endif - return d; - } - d[0] = 2; - d += 2; - } - #endif - - if (c3 >= matchMinPos && cur[(ptrdiff_t)c3 - (ptrdiff_t)m] == cur[0]) - { - d[1] = m - c3 - 1; - if (cur[(ptrdiff_t)c3 - (ptrdiff_t)m + 3] == cur[3]) - { - d[0] = 4; - return d + 2; - } - d[0] = 3; - d += 2; - } - - #ifdef _USE_H4 - if (c4 >= matchMinPos) - if ( - cur[(ptrdiff_t)c4 - (ptrdiff_t)m] == cur[0] && - cur[(ptrdiff_t)c4 - (ptrdiff_t)m + 3] == cur[3] - ) - { - *d++ = 4; - *d++ = m - c4 - 1; - } - #endif - - return d; -} - - -static UInt32* MatchFinderMt2_GetMatches(CMatchFinderMt *p, UInt32 *d) -{ - const UInt32 *bt = p->btBufPos; - const UInt32 len = *bt++; - const UInt32 *btLim = bt + len; - p->btBufPos = btLim; - p->btNumAvailBytes--; - INCREASE_LZ_POS - { - while (bt != btLim) - { - const UInt32 v0 = bt[0]; - const UInt32 v1 = bt[1]; - bt += 2; - d[0] = v0; - d[1] = v1; - d += 2; - } - } - return d; -} - - - -static UInt32* MatchFinderMt_GetMatches(CMatchFinderMt *p, UInt32 *d) -{ - const UInt32 *bt = p->btBufPos; - UInt32 len = *bt++; - const UInt32 avail = p->btNumAvailBytes - 1; - p->btNumAvailBytes = avail; - p->btBufPos = bt + len; - if (len == 0) - { - #define BT_HASH_BYTES_MAX 5 - if (avail >= (BT_HASH_BYTES_MAX - 1) - 1) - { - UInt32 m = p->lzPos; - if (m > p->historySize) - m -= p->historySize; - else - m = 1; - d = p->MixMatchesFunc(p, m, d); - } - } - else - { - /* - first match pair from BinTree: (match_len, match_dist), - (match_len >= numHashBytes). - MixMatchesFunc() inserts only hash matches that are nearer than (match_dist) - */ - d = p->MixMatchesFunc(p, p->lzPos - bt[1], d); - // if (d) // check for failure - do - { - const UInt32 v0 = bt[0]; - const UInt32 v1 = bt[1]; - bt += 2; - d[0] = v0; - d[1] = v1; - d += 2; - } - while (len -= 2); - } - INCREASE_LZ_POS - return d; -} - -#define SKIP_HEADER2_MT do { GET_NEXT_BLOCK_IF_REQUIRED -#define SKIP_HEADER_MT(n) SKIP_HEADER2_MT if (p->btNumAvailBytes-- >= (n)) { const Byte *cur = p->pointerToCurPos; UInt32 *hash = p->hash; -#define SKIP_FOOTER_MT } INCREASE_LZ_POS p->btBufPos += (size_t)*p->btBufPos + 1; } while (--num != 0); - -static void MatchFinderMt0_Skip(CMatchFinderMt *p, UInt32 num) -{ - SKIP_HEADER2_MT { p->btNumAvailBytes--; - SKIP_FOOTER_MT -} - -static void MatchFinderMt2_Skip(CMatchFinderMt *p, UInt32 num) -{ - SKIP_HEADER_MT(2) - UInt32 h2; - MT_HASH2_CALC - hash[h2] = p->lzPos; - SKIP_FOOTER_MT -} - -static void MatchFinderMt3_Skip(CMatchFinderMt *p, UInt32 num) -{ - SKIP_HEADER_MT(3) - UInt32 h2, h3; - MT_HASH3_CALC - (hash + kFix3HashSize)[h3] = - hash[ h2] = - p->lzPos; - SKIP_FOOTER_MT -} - -/* -// MatchFinderMt4_Skip() is similar to MatchFinderMt3_Skip(). -// The difference is that MatchFinderMt3_Skip() updates hash for last 3 bytes of stream. - -static void MatchFinderMt4_Skip(CMatchFinderMt *p, UInt32 num) -{ - SKIP_HEADER_MT(4) - UInt32 h2, h3; // h4 - MT_HASH3_CALC - // MT_HASH4_CALC - // (hash + kFix4HashSize)[h4] = - (hash + kFix3HashSize)[h3] = - hash[ h2] = - p->lzPos; - SKIP_FOOTER_MT -} -*/ - -void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder2 *vTable) -{ - vTable->Init = (Mf_Init_Func)MatchFinderMt_Init; - vTable->GetNumAvailableBytes = (Mf_GetNumAvailableBytes_Func)MatchFinderMt_GetNumAvailableBytes; - vTable->GetPointerToCurrentPos = (Mf_GetPointerToCurrentPos_Func)MatchFinderMt_GetPointerToCurrentPos; - vTable->GetMatches = (Mf_GetMatches_Func)MatchFinderMt_GetMatches; - - switch (MF(p)->numHashBytes) - { - case 2: - p->GetHeadsFunc = GetHeads2; - p->MixMatchesFunc = (Mf_Mix_Matches)NULL; - vTable->Skip = (Mf_Skip_Func)MatchFinderMt0_Skip; - vTable->GetMatches = (Mf_GetMatches_Func)MatchFinderMt2_GetMatches; - break; - case 3: - p->GetHeadsFunc = MF(p)->bigHash ? GetHeads3b : GetHeads3; - p->MixMatchesFunc = (Mf_Mix_Matches)MixMatches2; - vTable->Skip = (Mf_Skip_Func)MatchFinderMt2_Skip; - break; - case 4: - p->GetHeadsFunc = MF(p)->bigHash ? GetHeads4b : GetHeads4; - - // it's fast inline version of GetMatches() - // vTable->GetMatches = (Mf_GetMatches_Func)MatchFinderMt_GetMatches_Bt4; - - p->MixMatchesFunc = (Mf_Mix_Matches)MixMatches3; - vTable->Skip = (Mf_Skip_Func)MatchFinderMt3_Skip; - break; - default: - p->GetHeadsFunc = MF(p)->bigHash ? GetHeads5b : GetHeads5; - p->MixMatchesFunc = (Mf_Mix_Matches)MixMatches4; - vTable->Skip = - (Mf_Skip_Func)MatchFinderMt3_Skip; - // (Mf_Skip_Func)MatchFinderMt4_Skip; - break; - } -} +/* LzFindMt.c -- multithreaded Match finder for LZ algorithms +2021-12-21 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +// #include + +#include "CpuArch.h" + +#include "LzHash.h" +#include "LzFindMt.h" + +// #define LOG_ITERS + +// #define LOG_THREAD + +#ifdef LOG_THREAD +#include +#define PRF(x) x +#else +#define PRF(x) +#endif + +#ifdef LOG_ITERS +#include +extern UInt64 g_NumIters_Tree; +extern UInt64 g_NumIters_Loop; +extern UInt64 g_NumIters_Bytes; +#define LOG_ITER(x) x +#else +#define LOG_ITER(x) +#endif + +#define kMtHashBlockSize ((UInt32)1 << 17) +#define kMtHashNumBlocks (1 << 1) + +#define GET_HASH_BLOCK_OFFSET(i) (((i) & (kMtHashNumBlocks - 1)) * kMtHashBlockSize) + +#define kMtBtBlockSize ((UInt32)1 << 16) +#define kMtBtNumBlocks (1 << 4) + +#define GET_BT_BLOCK_OFFSET(i) (((i) & (kMtBtNumBlocks - 1)) * (size_t)kMtBtBlockSize) + +/* + HASH functions: + We use raw 8/16 bits from a[1] and a[2], + xored with crc(a[0]) and crc(a[3]). + We check a[0], a[3] only. We don't need to compare a[1] and a[2] in matches. + our crc() function provides one-to-one correspondence for low 8-bit values: + (crc[0...0xFF] & 0xFF) <-> [0...0xFF] +*/ + +#define MF(mt) ((mt)->MatchFinder) +#define MF_CRC (p->crc) + +// #define MF(mt) (&(mt)->MatchFinder) +// #define MF_CRC (p->MatchFinder.crc) + +#define MT_HASH2_CALC \ + h2 = (MF_CRC[cur[0]] ^ cur[1]) & (kHash2Size - 1); + +#define MT_HASH3_CALC { \ + UInt32 temp = MF_CRC[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + h3 = (temp ^ ((UInt32)cur[2] << 8)) & (kHash3Size - 1); } + +/* +#define MT_HASH3_CALC__NO_2 { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h3 = (temp ^ ((UInt32)cur[2] << 8)) & (kHash3Size - 1); } + +#define __MT_HASH4_CALC { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + temp ^= ((UInt32)cur[2] << 8); \ + h3 = temp & (kHash3Size - 1); \ + h4 = (temp ^ (p->crc[cur[3]] << kLzHash_CrcShift_1)) & p->hash4Mask; } + // (kHash4Size - 1); +*/ + + +MY_NO_INLINE +static void MtSync_Construct(CMtSync *p) +{ + p->affinity = 0; + p->wasCreated = False; + p->csWasInitialized = False; + p->csWasEntered = False; + Thread_Construct(&p->thread); + Event_Construct(&p->canStart); + Event_Construct(&p->wasStopped); + Semaphore_Construct(&p->freeSemaphore); + Semaphore_Construct(&p->filledSemaphore); +} + + +#define DEBUG_BUFFER_LOCK // define it to debug lock state + +#ifdef DEBUG_BUFFER_LOCK +#include +#define BUFFER_MUST_BE_LOCKED(p) if (!(p)->csWasEntered) exit(1); +#define BUFFER_MUST_BE_UNLOCKED(p) if ( (p)->csWasEntered) exit(1); +#else +#define BUFFER_MUST_BE_LOCKED(p) +#define BUFFER_MUST_BE_UNLOCKED(p) +#endif + +#define LOCK_BUFFER(p) { \ + BUFFER_MUST_BE_UNLOCKED(p); \ + CriticalSection_Enter(&(p)->cs); \ + (p)->csWasEntered = True; } + +#define UNLOCK_BUFFER(p) { \ + BUFFER_MUST_BE_LOCKED(p); \ + CriticalSection_Leave(&(p)->cs); \ + (p)->csWasEntered = False; } + + +MY_NO_INLINE +static UInt32 MtSync_GetNextBlock(CMtSync *p) +{ + UInt32 numBlocks = 0; + if (p->needStart) + { + BUFFER_MUST_BE_UNLOCKED(p) + p->numProcessedBlocks = 1; + p->needStart = False; + p->stopWriting = False; + p->exit = False; + Event_Reset(&p->wasStopped); + Event_Set(&p->canStart); + } + else + { + UNLOCK_BUFFER(p) + // we free current block + numBlocks = p->numProcessedBlocks++; + Semaphore_Release1(&p->freeSemaphore); + } + + // buffer is UNLOCKED here + Semaphore_Wait(&p->filledSemaphore); + LOCK_BUFFER(p); + return numBlocks; +} + + +/* if Writing (Processing) thread was started, we must call MtSync_StopWriting() */ + +MY_NO_INLINE +static void MtSync_StopWriting(CMtSync *p) +{ + if (!Thread_WasCreated(&p->thread) || p->needStart) + return; + + PRF(printf("\nMtSync_StopWriting %p\n", p)); + + if (p->csWasEntered) + { + /* we don't use buffer in this thread after StopWriting(). + So we UNLOCK buffer. + And we restore default UNLOCKED state for stopped thread */ + UNLOCK_BUFFER(p) + } + + /* We send (p->stopWriting) message and release freeSemaphore + to free current block. + So the thread will see (p->stopWriting) at some + iteration after Wait(freeSemaphore). + The thread doesn't need to fill all avail free blocks, + so we can get fast thread stop. + */ + + p->stopWriting = True; + Semaphore_Release1(&p->freeSemaphore); // check semaphore count !!! + + PRF(printf("\nMtSync_StopWriting %p : Event_Wait(&p->wasStopped)\n", p)); + Event_Wait(&p->wasStopped); + PRF(printf("\nMtSync_StopWriting %p : Event_Wait() finsihed\n", p)); + + /* 21.03 : we don't restore samaphore counters here. + We will recreate and reinit samaphores in next start */ + + p->needStart = True; +} + + +MY_NO_INLINE +static void MtSync_Destruct(CMtSync *p) +{ + PRF(printf("\nMtSync_Destruct %p\n", p)); + + if (Thread_WasCreated(&p->thread)) + { + /* we want thread to be in Stopped state before sending EXIT command. + note: stop(btSync) will stop (htSync) also */ + MtSync_StopWriting(p); + /* thread in Stopped state here : (p->needStart == true) */ + p->exit = True; + // if (p->needStart) // it's (true) + Event_Set(&p->canStart); // we send EXIT command to thread + Thread_Wait_Close(&p->thread); // we wait thread finishing + } + + if (p->csWasInitialized) + { + CriticalSection_Delete(&p->cs); + p->csWasInitialized = False; + } + p->csWasEntered = False; + + Event_Close(&p->canStart); + 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; } +// we want to get real system error codes here instead of SZ_ERROR_THREAD +#define RINOK_THREAD(x) RINOK(x) + + +// call it before each new file (when new starting is required): +MY_NO_INLINE +static SRes MtSync_Init(CMtSync *p, UInt32 numBlocks) +{ + WRes wres; + // BUFFER_MUST_BE_UNLOCKED(p) + if (!p->needStart || p->csWasEntered) + return SZ_ERROR_FAIL; + wres = Semaphore_OptCreateInit(&p->freeSemaphore, numBlocks, numBlocks); + if (wres == 0) + wres = Semaphore_OptCreateInit(&p->filledSemaphore, 0, numBlocks); + return MY_SRes_HRESULT_FROM_WRes(wres); +} + + +static WRes MtSync_Create_WRes(CMtSync *p, THREAD_FUNC_TYPE startAddress, void *obj) +{ + WRes wres; + + if (p->wasCreated) + return SZ_OK; + + RINOK_THREAD(CriticalSection_Init(&p->cs)); + p->csWasInitialized = True; + p->csWasEntered = False; + + RINOK_THREAD(AutoResetEvent_CreateNotSignaled(&p->canStart)); + RINOK_THREAD(AutoResetEvent_CreateNotSignaled(&p->wasStopped)); + + p->needStart = True; + p->exit = True; /* p->exit is unused before (canStart) Event. + But in case of some unexpected code failure we will get fast exit from thread */ + + // return ERROR_TOO_MANY_POSTS; // for debug + // return EINVAL; // for debug + + if (p->affinity != 0) + wres = Thread_Create_With_Affinity(&p->thread, startAddress, obj, (CAffinityMask)p->affinity); + else + wres = Thread_Create(&p->thread, startAddress, obj); + + RINOK_THREAD(wres); + p->wasCreated = True; + return SZ_OK; +} + + +MY_NO_INLINE +static SRes MtSync_Create(CMtSync *p, THREAD_FUNC_TYPE startAddress, void *obj) +{ + const WRes wres = MtSync_Create_WRes(p, startAddress, obj); + if (wres == 0) + return 0; + MtSync_Destruct(p); + return MY_SRes_HRESULT_FROM_WRes(wres); +} + + +// ---------- HASH THREAD ---------- + +#define kMtMaxValForNormalize 0xFFFFFFFF +// #define kMtMaxValForNormalize ((1 << 21)) // for debug +// #define kNormalizeAlign (1 << 7) // alignment for speculated accesses + +#ifdef MY_CPU_LE_UNALIGN + #define GetUi24hi_from32(p) ((UInt32)GetUi32(p) >> 8) +#else + #define GetUi24hi_from32(p) ((p)[1] ^ ((UInt32)(p)[2] << 8) ^ ((UInt32)(p)[3] << 16)) +#endif + +#define GetHeads_DECL(name) \ + static void GetHeads ## name(const Byte *p, UInt32 pos, \ + UInt32 *hash, UInt32 hashMask, UInt32 *heads, UInt32 numHeads, const UInt32 *crc) + +#define GetHeads_LOOP(v) \ + for (; numHeads != 0; numHeads--) { \ + const UInt32 value = (v); \ + p++; \ + *heads++ = pos - hash[value]; \ + hash[value] = pos++; } + +#define DEF_GetHeads2(name, v, action) \ + GetHeads_DECL(name) { action \ + GetHeads_LOOP(v) } + +#define DEF_GetHeads(name, v) DEF_GetHeads2(name, v, ;) + +DEF_GetHeads2(2, GetUi16(p), UNUSED_VAR(hashMask); UNUSED_VAR(crc); ) +DEF_GetHeads(3, (crc[p[0]] ^ GetUi16(p + 1)) & hashMask) +DEF_GetHeads2(3b, GetUi16(p) ^ ((UInt32)(p)[2] << 16), UNUSED_VAR(hashMask); UNUSED_VAR(crc); ) +// BT3 is not good for crc collisions for big hashMask values. + +/* +GetHeads_DECL(3b) +{ + UNUSED_VAR(hashMask); + UNUSED_VAR(crc); + { + const Byte *pLim = p + numHeads; + if (numHeads == 0) + return; + pLim--; + while (p < pLim) + { + UInt32 v1 = GetUi32(p); + UInt32 v0 = v1 & 0xFFFFFF; + UInt32 h0, h1; + p += 2; + v1 >>= 8; + h0 = hash[v0]; hash[v0] = pos; heads[0] = pos - h0; pos++; + h1 = hash[v1]; hash[v1] = pos; heads[1] = pos - h1; pos++; + heads += 2; + } + if (p == pLim) + { + UInt32 v0 = GetUi16(p) ^ ((UInt32)(p)[2] << 16); + *heads = pos - hash[v0]; + hash[v0] = pos; + } + } +} +*/ + +/* +GetHeads_DECL(4) +{ + unsigned sh = 0; + UNUSED_VAR(crc) + while ((hashMask & 0x80000000) == 0) + { + hashMask <<= 1; + sh++; + } + GetHeads_LOOP((GetUi32(p) * 0xa54a1) >> sh) +} +#define GetHeads4b GetHeads4 +*/ + +#define USE_GetHeads_LOCAL_CRC + +#ifdef USE_GetHeads_LOCAL_CRC + +GetHeads_DECL(4) +{ + UInt32 crc0[256]; + UInt32 crc1[256]; + { + unsigned i; + for (i = 0; i < 256; i++) + { + UInt32 v = crc[i]; + crc0[i] = v & hashMask; + crc1[i] = (v << kLzHash_CrcShift_1) & hashMask; + // crc1[i] = rotlFixed(v, 8) & hashMask; + } + } + GetHeads_LOOP(crc0[p[0]] ^ crc1[p[3]] ^ (UInt32)GetUi16(p+1)) +} + +GetHeads_DECL(4b) +{ + UInt32 crc0[256]; + { + unsigned i; + for (i = 0; i < 256; i++) + crc0[i] = crc[i] & hashMask; + } + GetHeads_LOOP(crc0[p[0]] ^ GetUi24hi_from32(p)) +} + +GetHeads_DECL(5) +{ + UInt32 crc0[256]; + UInt32 crc1[256]; + UInt32 crc2[256]; + { + unsigned i; + for (i = 0; i < 256; i++) + { + UInt32 v = crc[i]; + crc0[i] = v & hashMask; + crc1[i] = (v << kLzHash_CrcShift_1) & hashMask; + crc2[i] = (v << kLzHash_CrcShift_2) & hashMask; + } + } + GetHeads_LOOP(crc0[p[0]] ^ crc1[p[3]] ^ crc2[p[4]] ^ (UInt32)GetUi16(p+1)) +} + +GetHeads_DECL(5b) +{ + UInt32 crc0[256]; + UInt32 crc1[256]; + { + unsigned i; + for (i = 0; i < 256; i++) + { + UInt32 v = crc[i]; + crc0[i] = v & hashMask; + crc1[i] = (v << kLzHash_CrcShift_1) & hashMask; + } + } + GetHeads_LOOP(crc0[p[0]] ^ crc1[p[4]] ^ GetUi24hi_from32(p)) +} + +#else + +DEF_GetHeads(4, (crc[p[0]] ^ (crc[p[3]] << kLzHash_CrcShift_1) ^ (UInt32)GetUi16(p+1)) & hashMask) +DEF_GetHeads(4b, (crc[p[0]] ^ GetUi24hi_from32(p)) & hashMask) +DEF_GetHeads(5, (crc[p[0]] ^ (crc[p[3]] << kLzHash_CrcShift_1) ^ (crc[p[4]] << kLzHash_CrcShift_2) ^ (UInt32)GetUi16(p + 1)) & hashMask) +DEF_GetHeads(5b, (crc[p[0]] ^ (crc[p[4]] << kLzHash_CrcShift_1) ^ GetUi24hi_from32(p)) & hashMask) + +#endif + + +static void HashThreadFunc(CMatchFinderMt *mt) +{ + CMtSync *p = &mt->hashSync; + PRF(printf("\nHashThreadFunc\n")); + + for (;;) + { + UInt32 blockIndex = 0; + PRF(printf("\nHashThreadFunc : Event_Wait(&p->canStart)\n")); + Event_Wait(&p->canStart); + PRF(printf("\nHashThreadFunc : Event_Wait(&p->canStart) : after \n")); + if (p->exit) + { + PRF(printf("\nHashThreadFunc : exit \n")); + return; + } + + MatchFinder_Init_HighHash(MF(mt)); + + for (;;) + { + PRF(printf("Hash thread block = %d pos = %d\n", (unsigned)blockIndex, mt->MatchFinder->pos)); + + { + CMatchFinder *mf = MF(mt); + if (MatchFinder_NeedMove(mf)) + { + CriticalSection_Enter(&mt->btSync.cs); + CriticalSection_Enter(&mt->hashSync.cs); + { + const Byte *beforePtr = Inline_MatchFinder_GetPointerToCurrentPos(mf); + ptrdiff_t offset; + MatchFinder_MoveBlock(mf); + offset = beforePtr - Inline_MatchFinder_GetPointerToCurrentPos(mf); + mt->pointerToCurPos -= offset; + mt->buffer -= offset; + } + CriticalSection_Leave(&mt->hashSync.cs); + CriticalSection_Leave(&mt->btSync.cs); + continue; + } + + Semaphore_Wait(&p->freeSemaphore); + + if (p->exit) // exit is unexpected here. But we check it here for some failure case + return; + + // for faster stop : we check (p->stopWriting) after Wait(freeSemaphore) + if (p->stopWriting) + break; + + MatchFinder_ReadIfRequired(mf); + { + UInt32 *heads = mt->hashBuf + GET_HASH_BLOCK_OFFSET(blockIndex++); + UInt32 num = Inline_MatchFinder_GetNumAvailableBytes(mf); + heads[0] = 2; + heads[1] = num; + + /* heads[1] contains the number of avail bytes: + if (avail < mf->numHashBytes) : + { + it means that stream was finished + HASH_THREAD and BT_TREAD must move position for heads[1] (avail) bytes. + HASH_THREAD doesn't stop, + HASH_THREAD fills only the header (2 numbers) for all next blocks: + {2, NumHashBytes - 1}, {2,0}, {2,0}, ... , {2,0} + } + else + { + HASH_THREAD and BT_TREAD must move position for (heads[0] - 2) bytes; + } + */ + + if (num >= mf->numHashBytes) + { + num = num - mf->numHashBytes + 1; + if (num > kMtHashBlockSize - 2) + num = kMtHashBlockSize - 2; + + if (mf->pos > (UInt32)kMtMaxValForNormalize - num) + { + const UInt32 subValue = (mf->pos - mf->historySize - 1); // & ~(UInt32)(kNormalizeAlign - 1); + Inline_MatchFinder_ReduceOffsets(mf, subValue); + MatchFinder_Normalize3(subValue, mf->hash + mf->fixedHashSize, (size_t)mf->hashMask + 1); + } + + heads[0] = 2 + num; + mt->GetHeadsFunc(mf->buffer, mf->pos, mf->hash + mf->fixedHashSize, mf->hashMask, heads + 2, num, mf->crc); + } + + mf->pos += num; // wrap over zero is allowed at the end of stream + mf->buffer += num; + } + } + + Semaphore_Release1(&p->filledSemaphore); + } // for() processing end + + // p->numBlocks_Sent = blockIndex; + Event_Set(&p->wasStopped); + } // for() thread end +} + + + + +// ---------- BT THREAD ---------- + +/* we use one variable instead of two (cyclicBufferPos == pos) before CyclicBuf wrap. + here we define fixed offset of (p->pos) from (p->cyclicBufferPos) */ +#define CYC_TO_POS_OFFSET 0 +// #define CYC_TO_POS_OFFSET 1 // for debug + +#define MFMT_GM_INLINE + +#ifdef MFMT_GM_INLINE + +/* + we use size_t for (pos) instead of UInt32 + to eliminate "movsx" BUG in old MSVC x64 compiler. +*/ + + +UInt32 * MY_FAST_CALL GetMatchesSpecN_2(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, + UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, + UInt32 *posRes); + +#endif + + +static void BtGetMatches(CMatchFinderMt *p, UInt32 *d) +{ + UInt32 numProcessed = 0; + UInt32 curPos = 2; + + /* GetMatchesSpec() functions don't create (len = 1) + in [len, dist] match pairs, if (p->numHashBytes >= 2) + Also we suppose here that (matchMaxLen >= 2). + So the following code for (reserve) is not required + UInt32 reserve = (p->matchMaxLen * 2); + const UInt32 kNumHashBytes_Max = 5; // BT_HASH_BYTES_MAX + if (reserve < kNumHashBytes_Max - 1) + reserve = kNumHashBytes_Max - 1; + const UInt32 limit = kMtBtBlockSize - (reserve); + */ + + const UInt32 limit = kMtBtBlockSize - (p->matchMaxLen * 2); + + d[1] = p->hashNumAvail; + + if (p->failure_BT) + { + // printf("\n == 1 BtGetMatches() p->failure_BT\n"); + d[0] = 0; + // d[1] = 0; + return; + } + + while (curPos < limit) + { + if (p->hashBufPos == p->hashBufPosLimit) + { + // MatchFinderMt_GetNextBlock_Hash(p); + UInt32 avail; + { + const UInt32 bi = MtSync_GetNextBlock(&p->hashSync); + const UInt32 k = GET_HASH_BLOCK_OFFSET(bi); + const UInt32 *h = p->hashBuf + k; + avail = h[1]; + p->hashBufPosLimit = k + h[0]; + p->hashNumAvail = avail; + p->hashBufPos = k + 2; + } + + { + /* we must prevent UInt32 overflow for avail total value, + if avail was increased with new hash block */ + UInt32 availSum = numProcessed + avail; + if (availSum < numProcessed) + availSum = (UInt32)(Int32)-1; + d[1] = availSum; + } + + if (avail >= p->numHashBytes) + continue; + + // if (p->hashBufPos != p->hashBufPosLimit) exit(1); + + /* (avail < p->numHashBytes) + It means that stream was finished. + And (avail) - is a number of remaining bytes, + we fill (d) for (avail) bytes for LZ_THREAD (receiver). + but we don't update (p->pos) and (p->cyclicBufferPos) here in BT_THREAD */ + + /* here we suppose that we have space enough: + (kMtBtBlockSize - curPos >= p->hashNumAvail) */ + p->hashNumAvail = 0; + d[0] = curPos + avail; + d += curPos; + for (; avail != 0; avail--) + *d++ = 0; + return; + } + { + UInt32 size = p->hashBufPosLimit - p->hashBufPos; + UInt32 pos = p->pos; + UInt32 cyclicBufferPos = p->cyclicBufferPos; + UInt32 lenLimit = p->matchMaxLen; + 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; + } + + if (pos > (UInt32)kMtMaxValForNormalize - size) + { + const UInt32 subValue = (pos - p->cyclicBufferSize); // & ~(UInt32)(kNormalizeAlign - 1); + pos -= subValue; + p->pos = pos; + MatchFinder_Normalize3(subValue, p->son, (size_t)p->cyclicBufferSize * 2); + } + + #ifndef MFMT_GM_INLINE + while (curPos < limit && size-- != 0) + { + UInt32 *startDistances = d + 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 = pos; + const UInt32 *d_end; + { + d_end = GetMatchesSpecN_2( + p->buffer + lenLimit - 1, + pos, p->buffer, p->son, p->cutValue, d + curPos, + p->numHashBytes - 1, p->hashBuf + p->hashBufPos, + d + limit, p->hashBuf + p->hashBufPos + size, + cyclicBufferPos, p->cyclicBufferSize, + &posRes); + } + { + if (!d_end) + { + // printf("\n == 2 BtGetMatches() p->failure_BT\n"); + // internal data failure + p->failure_BT = True; + d[0] = 0; + // d[1] = 0; + return; + } + } + curPos = (UInt32)(d_end - d); + { + const UInt32 processed = posRes - pos; + pos = posRes; + p->hashBufPos += processed; + cyclicBufferPos += processed; + p->buffer += processed; + } + } + #endif + + { + const UInt32 processed = pos - p->pos; + numProcessed += processed; + p->hashNumAvail -= processed; + p->pos = pos; + } + if (cyclicBufferPos == p->cyclicBufferSize) + cyclicBufferPos = 0; + p->cyclicBufferPos = cyclicBufferPos; + } + } + + d[0] = curPos; +} + + +static void BtFillBlock(CMatchFinderMt *p, UInt32 globalBlockIndex) +{ + CMtSync *sync = &p->hashSync; + + BUFFER_MUST_BE_UNLOCKED(sync) + + if (!sync->needStart) + { + LOCK_BUFFER(sync) + } + + BtGetMatches(p, p->btBuf + GET_BT_BLOCK_OFFSET(globalBlockIndex)); + + /* We suppose that we have called GetNextBlock() from start. + So buffer is LOCKED */ + + UNLOCK_BUFFER(sync) +} + + +MY_NO_INLINE +static void BtThreadFunc(CMatchFinderMt *mt) +{ + CMtSync *p = &mt->btSync; + for (;;) + { + UInt32 blockIndex = 0; + Event_Wait(&p->canStart); + + for (;;) + { + PRF(printf(" BT thread block = %d pos = %d\n", (unsigned)blockIndex, mt->pos)); + /* (p->exit == true) is possible after (p->canStart) at first loop iteration + and is unexpected after more Wait(freeSemaphore) iterations */ + if (p->exit) + return; + + Semaphore_Wait(&p->freeSemaphore); + + // for faster stop : we check (p->stopWriting) after Wait(freeSemaphore) + if (p->stopWriting) + break; + + BtFillBlock(mt, blockIndex++); + + Semaphore_Release1(&p->filledSemaphore); + } + + // we stop HASH_THREAD here + MtSync_StopWriting(&mt->hashSync); + + // p->numBlocks_Sent = blockIndex; + Event_Set(&p->wasStopped); + } +} + + +void MatchFinderMt_Construct(CMatchFinderMt *p) +{ + p->hashBuf = NULL; + MtSync_Construct(&p->hashSync); + MtSync_Construct(&p->btSync); +} + +static void MatchFinderMt_FreeMem(CMatchFinderMt *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->hashBuf); + p->hashBuf = NULL; +} + +void MatchFinderMt_Destruct(CMatchFinderMt *p, ISzAllocPtr alloc) +{ + /* + HASH_THREAD can use CriticalSection(s) btSync.cs and hashSync.cs. + So we must be sure that HASH_THREAD will not use CriticalSection(s) + after deleting CriticalSection here. + + we call ReleaseStream(p) + that calls StopWriting(btSync) + that calls StopWriting(hashSync), if it's required to stop HASH_THREAD. + after StopWriting() it's safe to destruct MtSync(s) in any order */ + + MatchFinderMt_ReleaseStream(p); + + MtSync_Destruct(&p->btSync); + MtSync_Destruct(&p->hashSync); + + LOG_ITER( + printf("\nTree %9d * %7d iter = %9d = sum : bytes = %9d\n", + (UInt32)(g_NumIters_Tree / 1000), + (UInt32)(((UInt64)g_NumIters_Loop * 1000) / (g_NumIters_Tree + 1)), + (UInt32)(g_NumIters_Loop / 1000), + (UInt32)(g_NumIters_Bytes / 1000) + )); + + MatchFinderMt_FreeMem(p, alloc); +} + + +#define kHashBufferSize (kMtHashBlockSize * kMtHashNumBlocks) +#define kBtBufferSize (kMtBtBlockSize * kMtBtNumBlocks) + + +static THREAD_FUNC_DECL HashThreadFunc2(void *p) { HashThreadFunc((CMatchFinderMt *)p); return 0; } +static THREAD_FUNC_DECL BtThreadFunc2(void *p) +{ + Byte allocaDummy[0x180]; + unsigned i = 0; + for (i = 0; i < 16; i++) + allocaDummy[i] = (Byte)0; + if (allocaDummy[0] == 0) + BtThreadFunc((CMatchFinderMt *)p); + return 0; +} + + +SRes MatchFinderMt_Create(CMatchFinderMt *p, UInt32 historySize, UInt32 keepAddBufferBefore, + UInt32 matchMaxLen, UInt32 keepAddBufferAfter, ISzAllocPtr alloc) +{ + CMatchFinder *mf = MF(p); + p->historySize = historySize; + if (kMtBtBlockSize <= matchMaxLen * 4) + return SZ_ERROR_PARAM; + if (!p->hashBuf) + { + p->hashBuf = (UInt32 *)ISzAlloc_Alloc(alloc, ((size_t)kHashBufferSize + (size_t)kBtBufferSize) * sizeof(UInt32)); + if (!p->hashBuf) + return SZ_ERROR_MEM; + p->btBuf = p->hashBuf + kHashBufferSize; + } + keepAddBufferBefore += (kHashBufferSize + kBtBufferSize); + keepAddBufferAfter += kMtHashBlockSize; + if (!MatchFinder_Create(mf, historySize, keepAddBufferBefore, matchMaxLen, keepAddBufferAfter, alloc)) + return SZ_ERROR_MEM; + + RINOK(MtSync_Create(&p->hashSync, HashThreadFunc2, p)); + RINOK(MtSync_Create(&p->btSync, BtThreadFunc2, p)); + return SZ_OK; +} + + +SRes MatchFinderMt_InitMt(CMatchFinderMt *p) +{ + RINOK(MtSync_Init(&p->hashSync, kMtHashNumBlocks)); + return MtSync_Init(&p->btSync, kMtBtNumBlocks); +} + + +static void MatchFinderMt_Init(CMatchFinderMt *p) +{ + CMatchFinder *mf = MF(p); + + p->btBufPos = + p->btBufPosLimit = NULL; + p->hashBufPos = + p->hashBufPosLimit = 0; + p->hashNumAvail = 0; // 21.03 + + p->failure_BT = False; + + /* Init without data reading. We don't want to read data in this thread */ + MatchFinder_Init_4(mf); + + MatchFinder_Init_LowHash(mf); + + p->pointerToCurPos = Inline_MatchFinder_GetPointerToCurrentPos(mf); + p->btNumAvailBytes = 0; + p->failure_LZ_BT = False; + // p->failure_LZ_LZ = False; + + p->lzPos = + 1; // optimal smallest value + // 0; // for debug: ignores match to start + // kNormalizeAlign; // for debug + + p->hash = mf->hash; + p->fixedHashSize = mf->fixedHashSize; + // p->hash4Mask = mf->hash4Mask; + p->crc = mf->crc; + // memcpy(p->crc, mf->crc, sizeof(mf->crc)); + + p->son = mf->son; + p->matchMaxLen = mf->matchMaxLen; + p->numHashBytes = mf->numHashBytes; + + /* (mf->pos) and (mf->streamPos) were already initialized to 1 in MatchFinder_Init_4() */ + // mf->streamPos = mf->pos = 1; // optimal smallest value + // 0; // for debug: ignores match to start + // kNormalizeAlign; // for debug + + /* we must init (p->pos = mf->pos) for BT, because + BT code needs (p->pos == delta_value_for_empty_hash_record == mf->pos) */ + p->pos = mf->pos; // do not change it + + p->cyclicBufferPos = (p->pos - CYC_TO_POS_OFFSET); + p->cyclicBufferSize = mf->cyclicBufferSize; + p->buffer = mf->buffer; + p->cutValue = mf->cutValue; + // p->son[0] = p->son[1] = 0; // unused: to init skipped record for speculated accesses. +} + + +/* ReleaseStream is required to finish multithreading */ +void MatchFinderMt_ReleaseStream(CMatchFinderMt *p) +{ + // Sleep(1); // for debug + MtSync_StopWriting(&p->btSync); + // Sleep(200); // for debug + /* p->MatchFinder->ReleaseStream(); */ +} + + +MY_NO_INLINE +static UInt32 MatchFinderMt_GetNextBlock_Bt(CMatchFinderMt *p) +{ + if (p->failure_LZ_BT) + p->btBufPos = p->failureBuf; + else + { + const UInt32 bi = MtSync_GetNextBlock(&p->btSync); + const UInt32 *bt = p->btBuf + GET_BT_BLOCK_OFFSET(bi); + { + const UInt32 numItems = bt[0]; + p->btBufPosLimit = bt + numItems; + p->btNumAvailBytes = bt[1]; + p->btBufPos = bt + 2; + if (numItems < 2 || numItems > kMtBtBlockSize) + { + p->failureBuf[0] = 0; + p->btBufPos = p->failureBuf; + p->btBufPosLimit = p->failureBuf + 1; + p->failure_LZ_BT = True; + // p->btNumAvailBytes = 0; + /* we don't want to decrease AvailBytes, that was load before. + that can be unxepected for the code that have loaded anopther value before */ + } + } + + if (p->lzPos >= (UInt32)kMtMaxValForNormalize - (UInt32)kMtBtBlockSize) + { + /* we don't check (lzPos) over exact avail bytes in (btBuf). + (fixedHashSize) is small, so normalization is fast */ + const UInt32 subValue = (p->lzPos - p->historySize - 1); // & ~(UInt32)(kNormalizeAlign - 1); + p->lzPos -= subValue; + MatchFinder_Normalize3(subValue, p->hash, p->fixedHashSize); + } + } + return p->btNumAvailBytes; +} + + + +static const Byte * MatchFinderMt_GetPointerToCurrentPos(CMatchFinderMt *p) +{ + return p->pointerToCurPos; +} + + +#define GET_NEXT_BLOCK_IF_REQUIRED if (p->btBufPos == p->btBufPosLimit) MatchFinderMt_GetNextBlock_Bt(p); + + +static UInt32 MatchFinderMt_GetNumAvailableBytes(CMatchFinderMt *p) +{ + if (p->btBufPos != p->btBufPosLimit) + return p->btNumAvailBytes; + return MatchFinderMt_GetNextBlock_Bt(p); +} + + +// #define CHECK_FAILURE_LZ(_match_, _pos_) if (_match_ >= _pos_) { p->failure_LZ_LZ = True; return d; } +#define CHECK_FAILURE_LZ(_match_, _pos_) + +static UInt32 * MixMatches2(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *d) +{ + UInt32 h2, c2; + UInt32 *hash = p->hash; + const Byte *cur = p->pointerToCurPos; + const UInt32 m = p->lzPos; + MT_HASH2_CALC + + c2 = hash[h2]; + hash[h2] = m; + + if (c2 >= matchMinPos) + { + CHECK_FAILURE_LZ(c2, m) + if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m] == cur[0]) + { + *d++ = 2; + *d++ = m - c2 - 1; + } + } + + return d; +} + +static UInt32 * MixMatches3(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *d) +{ + UInt32 h2, h3, c2, c3; + UInt32 *hash = p->hash; + const Byte *cur = p->pointerToCurPos; + const UInt32 m = p->lzPos; + MT_HASH3_CALC + + c2 = hash[h2]; + c3 = (hash + kFix3HashSize)[h3]; + + hash[h2] = m; + (hash + kFix3HashSize)[h3] = m; + + if (c2 >= matchMinPos) + { + CHECK_FAILURE_LZ(c2, m) + if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m] == cur[0]) + { + d[1] = m - c2 - 1; + if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m + 2] == cur[2]) + { + d[0] = 3; + return d + 2; + } + d[0] = 2; + d += 2; + } + } + + if (c3 >= matchMinPos) + { + CHECK_FAILURE_LZ(c3, m) + if (cur[(ptrdiff_t)c3 - (ptrdiff_t)m] == cur[0]) + { + *d++ = 3; + *d++ = m - c3 - 1; + } + } + + return d; +} + + +#define INCREASE_LZ_POS p->lzPos++; p->pointerToCurPos++; + +/* +static +UInt32* MatchFinderMt_GetMatches_Bt4(CMatchFinderMt *p, UInt32 *d) +{ + const UInt32 *bt = p->btBufPos; + const UInt32 len = *bt++; + const UInt32 *btLim = bt + len; + UInt32 matchMinPos; + UInt32 avail = p->btNumAvailBytes - 1; + p->btBufPos = btLim; + + { + p->btNumAvailBytes = avail; + + #define BT_HASH_BYTES_MAX 5 + + matchMinPos = p->lzPos; + + if (len != 0) + matchMinPos -= bt[1]; + else if (avail < (BT_HASH_BYTES_MAX - 1) - 1) + { + INCREASE_LZ_POS + return d; + } + else + { + const UInt32 hs = p->historySize; + if (matchMinPos > hs) + matchMinPos -= hs; + else + matchMinPos = 1; + } + } + + for (;;) + { + + UInt32 h2, h3, c2, c3; + UInt32 *hash = p->hash; + const Byte *cur = p->pointerToCurPos; + UInt32 m = p->lzPos; + MT_HASH3_CALC + + c2 = hash[h2]; + c3 = (hash + kFix3HashSize)[h3]; + + hash[h2] = m; + (hash + kFix3HashSize)[h3] = m; + + if (c2 >= matchMinPos && cur[(ptrdiff_t)c2 - (ptrdiff_t)m] == cur[0]) + { + d[1] = m - c2 - 1; + if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m + 2] == cur[2]) + { + d[0] = 3; + d += 2; + break; + } + // else + { + d[0] = 2; + d += 2; + } + } + if (c3 >= matchMinPos && cur[(ptrdiff_t)c3 - (ptrdiff_t)m] == cur[0]) + { + *d++ = 3; + *d++ = m - c3 - 1; + } + break; + } + + if (len != 0) + { + do + { + const UInt32 v0 = bt[0]; + const UInt32 v1 = bt[1]; + bt += 2; + d[0] = v0; + d[1] = v1; + d += 2; + } + while (bt != btLim); + } + INCREASE_LZ_POS + return d; +} +*/ + + +static UInt32 *MixMatches4(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *d) +{ + UInt32 h2, h3, /* h4, */ c2, c3 /* , c4 */; + UInt32 *hash = p->hash; + const Byte *cur = p->pointerToCurPos; + const UInt32 m = p->lzPos; + MT_HASH3_CALC + // MT_HASH4_CALC + c2 = hash[h2]; + c3 = (hash + kFix3HashSize)[h3]; + // c4 = (hash + kFix4HashSize)[h4]; + + hash[h2] = m; + (hash + kFix3HashSize)[h3] = m; + // (hash + kFix4HashSize)[h4] = m; + + #define _USE_H2 + + #ifdef _USE_H2 + if (c2 >= matchMinPos && cur[(ptrdiff_t)c2 - (ptrdiff_t)m] == cur[0]) + { + d[1] = m - c2 - 1; + if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m + 2] == cur[2]) + { + // d[0] = (cur[(ptrdiff_t)c2 - (ptrdiff_t)m + 3] == cur[3]) ? 4 : 3; + // return d + 2; + + if (cur[(ptrdiff_t)c2 - (ptrdiff_t)m + 3] == cur[3]) + { + d[0] = 4; + return d + 2; + } + d[0] = 3; + d += 2; + + #ifdef _USE_H4 + if (c4 >= matchMinPos) + if ( + cur[(ptrdiff_t)c4 - (ptrdiff_t)m] == cur[0] && + cur[(ptrdiff_t)c4 - (ptrdiff_t)m + 3] == cur[3] + ) + { + *d++ = 4; + *d++ = m - c4 - 1; + } + #endif + return d; + } + d[0] = 2; + d += 2; + } + #endif + + if (c3 >= matchMinPos && cur[(ptrdiff_t)c3 - (ptrdiff_t)m] == cur[0]) + { + d[1] = m - c3 - 1; + if (cur[(ptrdiff_t)c3 - (ptrdiff_t)m + 3] == cur[3]) + { + d[0] = 4; + return d + 2; + } + d[0] = 3; + d += 2; + } + + #ifdef _USE_H4 + if (c4 >= matchMinPos) + if ( + cur[(ptrdiff_t)c4 - (ptrdiff_t)m] == cur[0] && + cur[(ptrdiff_t)c4 - (ptrdiff_t)m + 3] == cur[3] + ) + { + *d++ = 4; + *d++ = m - c4 - 1; + } + #endif + + return d; +} + + +static UInt32* MatchFinderMt2_GetMatches(CMatchFinderMt *p, UInt32 *d) +{ + const UInt32 *bt = p->btBufPos; + const UInt32 len = *bt++; + const UInt32 *btLim = bt + len; + p->btBufPos = btLim; + p->btNumAvailBytes--; + INCREASE_LZ_POS + { + while (bt != btLim) + { + const UInt32 v0 = bt[0]; + const UInt32 v1 = bt[1]; + bt += 2; + d[0] = v0; + d[1] = v1; + d += 2; + } + } + return d; +} + + + +static UInt32* MatchFinderMt_GetMatches(CMatchFinderMt *p, UInt32 *d) +{ + const UInt32 *bt = p->btBufPos; + UInt32 len = *bt++; + const UInt32 avail = p->btNumAvailBytes - 1; + p->btNumAvailBytes = avail; + p->btBufPos = bt + len; + if (len == 0) + { + #define BT_HASH_BYTES_MAX 5 + if (avail >= (BT_HASH_BYTES_MAX - 1) - 1) + { + UInt32 m = p->lzPos; + if (m > p->historySize) + m -= p->historySize; + else + m = 1; + d = p->MixMatchesFunc(p, m, d); + } + } + else + { + /* + first match pair from BinTree: (match_len, match_dist), + (match_len >= numHashBytes). + MixMatchesFunc() inserts only hash matches that are nearer than (match_dist) + */ + d = p->MixMatchesFunc(p, p->lzPos - bt[1], d); + // if (d) // check for failure + do + { + const UInt32 v0 = bt[0]; + const UInt32 v1 = bt[1]; + bt += 2; + d[0] = v0; + d[1] = v1; + d += 2; + } + while (len -= 2); + } + INCREASE_LZ_POS + return d; +} + +#define SKIP_HEADER2_MT do { GET_NEXT_BLOCK_IF_REQUIRED +#define SKIP_HEADER_MT(n) SKIP_HEADER2_MT if (p->btNumAvailBytes-- >= (n)) { const Byte *cur = p->pointerToCurPos; UInt32 *hash = p->hash; +#define SKIP_FOOTER_MT } INCREASE_LZ_POS p->btBufPos += (size_t)*p->btBufPos + 1; } while (--num != 0); + +static void MatchFinderMt0_Skip(CMatchFinderMt *p, UInt32 num) +{ + SKIP_HEADER2_MT { p->btNumAvailBytes--; + SKIP_FOOTER_MT +} + +static void MatchFinderMt2_Skip(CMatchFinderMt *p, UInt32 num) +{ + SKIP_HEADER_MT(2) + UInt32 h2; + MT_HASH2_CALC + hash[h2] = p->lzPos; + SKIP_FOOTER_MT +} + +static void MatchFinderMt3_Skip(CMatchFinderMt *p, UInt32 num) +{ + SKIP_HEADER_MT(3) + UInt32 h2, h3; + MT_HASH3_CALC + (hash + kFix3HashSize)[h3] = + hash[ h2] = + p->lzPos; + SKIP_FOOTER_MT +} + +/* +// MatchFinderMt4_Skip() is similar to MatchFinderMt3_Skip(). +// The difference is that MatchFinderMt3_Skip() updates hash for last 3 bytes of stream. + +static void MatchFinderMt4_Skip(CMatchFinderMt *p, UInt32 num) +{ + SKIP_HEADER_MT(4) + UInt32 h2, h3; // h4 + MT_HASH3_CALC + // MT_HASH4_CALC + // (hash + kFix4HashSize)[h4] = + (hash + kFix3HashSize)[h3] = + hash[ h2] = + p->lzPos; + SKIP_FOOTER_MT +} +*/ + +void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder2 *vTable) +{ + vTable->Init = (Mf_Init_Func)MatchFinderMt_Init; + vTable->GetNumAvailableBytes = (Mf_GetNumAvailableBytes_Func)MatchFinderMt_GetNumAvailableBytes; + vTable->GetPointerToCurrentPos = (Mf_GetPointerToCurrentPos_Func)MatchFinderMt_GetPointerToCurrentPos; + vTable->GetMatches = (Mf_GetMatches_Func)MatchFinderMt_GetMatches; + + switch (MF(p)->numHashBytes) + { + case 2: + p->GetHeadsFunc = GetHeads2; + p->MixMatchesFunc = (Mf_Mix_Matches)NULL; + vTable->Skip = (Mf_Skip_Func)MatchFinderMt0_Skip; + vTable->GetMatches = (Mf_GetMatches_Func)MatchFinderMt2_GetMatches; + break; + case 3: + p->GetHeadsFunc = MF(p)->bigHash ? GetHeads3b : GetHeads3; + p->MixMatchesFunc = (Mf_Mix_Matches)MixMatches2; + vTable->Skip = (Mf_Skip_Func)MatchFinderMt2_Skip; + break; + case 4: + p->GetHeadsFunc = MF(p)->bigHash ? GetHeads4b : GetHeads4; + + // it's fast inline version of GetMatches() + // vTable->GetMatches = (Mf_GetMatches_Func)MatchFinderMt_GetMatches_Bt4; + + p->MixMatchesFunc = (Mf_Mix_Matches)MixMatches3; + vTable->Skip = (Mf_Skip_Func)MatchFinderMt3_Skip; + break; + default: + p->GetHeadsFunc = MF(p)->bigHash ? GetHeads5b : GetHeads5; + p->MixMatchesFunc = (Mf_Mix_Matches)MixMatches4; + vTable->Skip = + (Mf_Skip_Func)MatchFinderMt3_Skip; + // (Mf_Skip_Func)MatchFinderMt4_Skip; + break; + } +} diff --git a/deps/LZMA-SDK/C/LzFindMt.h b/deps/LZMA-SDK/C/LzFindMt.h index ee9a1b6fe..660b7244d 100644 --- a/deps/LZMA-SDK/C/LzFindMt.h +++ b/deps/LZMA-SDK/C/LzFindMt.h @@ -1,109 +1,109 @@ -/* LzFindMt.h -- multithreaded Match finder for LZ algorithms -2021-07-12 : Igor Pavlov : Public domain */ - -#ifndef __LZ_FIND_MT_H -#define __LZ_FIND_MT_H - -#include "LzFind.h" -#include "Threads.h" - -EXTERN_C_BEGIN - -typedef struct _CMtSync -{ - UInt32 numProcessedBlocks; - CThread thread; - UInt64 affinity; - - BoolInt wasCreated; - BoolInt needStart; - BoolInt csWasInitialized; - BoolInt csWasEntered; - - BoolInt exit; - BoolInt stopWriting; - - CAutoResetEvent canStart; - CAutoResetEvent wasStopped; - CSemaphore freeSemaphore; - CSemaphore filledSemaphore; - CCriticalSection cs; - // UInt32 numBlocks_Sent; -} CMtSync; - -typedef UInt32 * (*Mf_Mix_Matches)(void *p, UInt32 matchMinPos, UInt32 *distances); - -/* kMtCacheLineDummy must be >= size_of_CPU_cache_line */ -#define kMtCacheLineDummy 128 - -typedef void (*Mf_GetHeads)(const Byte *buffer, UInt32 pos, - UInt32 *hash, UInt32 hashMask, UInt32 *heads, UInt32 numHeads, const UInt32 *crc); - -typedef struct _CMatchFinderMt -{ - /* LZ */ - const Byte *pointerToCurPos; - UInt32 *btBuf; - const UInt32 *btBufPos; - const UInt32 *btBufPosLimit; - UInt32 lzPos; - UInt32 btNumAvailBytes; - - UInt32 *hash; - UInt32 fixedHashSize; - // UInt32 hash4Mask; - UInt32 historySize; - const UInt32 *crc; - - Mf_Mix_Matches MixMatchesFunc; - UInt32 failure_LZ_BT; // failure in BT transfered to LZ - // UInt32 failure_LZ_LZ; // failure in LZ tables - UInt32 failureBuf[1]; - // UInt32 crc[256]; - - /* LZ + BT */ - CMtSync btSync; - Byte btDummy[kMtCacheLineDummy]; - - /* BT */ - UInt32 *hashBuf; - UInt32 hashBufPos; - UInt32 hashBufPosLimit; - UInt32 hashNumAvail; - UInt32 failure_BT; - - - CLzRef *son; - UInt32 matchMaxLen; - UInt32 numHashBytes; - UInt32 pos; - const Byte *buffer; - UInt32 cyclicBufferPos; - UInt32 cyclicBufferSize; /* it must be = (historySize + 1) */ - UInt32 cutValue; - - /* BT + Hash */ - CMtSync hashSync; - /* Byte hashDummy[kMtCacheLineDummy]; */ - - /* Hash */ - Mf_GetHeads GetHeadsFunc; - CMatchFinder *MatchFinder; - // CMatchFinder MatchFinder; -} CMatchFinderMt; - -// only for Mt part -void MatchFinderMt_Construct(CMatchFinderMt *p); -void MatchFinderMt_Destruct(CMatchFinderMt *p, ISzAllocPtr alloc); - -SRes MatchFinderMt_Create(CMatchFinderMt *p, UInt32 historySize, UInt32 keepAddBufferBefore, - UInt32 matchMaxLen, UInt32 keepAddBufferAfter, ISzAllocPtr alloc); -void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder2 *vTable); - -/* call MatchFinderMt_InitMt() before IMatchFinder::Init() */ -SRes MatchFinderMt_InitMt(CMatchFinderMt *p); -void MatchFinderMt_ReleaseStream(CMatchFinderMt *p); - -EXTERN_C_END - -#endif +/* LzFindMt.h -- multithreaded Match finder for LZ algorithms +2021-07-12 : Igor Pavlov : Public domain */ + +#ifndef __LZ_FIND_MT_H +#define __LZ_FIND_MT_H + +#include "LzFind.h" +#include "Threads.h" + +EXTERN_C_BEGIN + +typedef struct _CMtSync +{ + UInt32 numProcessedBlocks; + CThread thread; + UInt64 affinity; + + BoolInt wasCreated; + BoolInt needStart; + BoolInt csWasInitialized; + BoolInt csWasEntered; + + BoolInt exit; + BoolInt stopWriting; + + CAutoResetEvent canStart; + CAutoResetEvent wasStopped; + CSemaphore freeSemaphore; + CSemaphore filledSemaphore; + CCriticalSection cs; + // UInt32 numBlocks_Sent; +} CMtSync; + +typedef UInt32 * (*Mf_Mix_Matches)(void *p, UInt32 matchMinPos, UInt32 *distances); + +/* kMtCacheLineDummy must be >= size_of_CPU_cache_line */ +#define kMtCacheLineDummy 128 + +typedef void (*Mf_GetHeads)(const Byte *buffer, UInt32 pos, + UInt32 *hash, UInt32 hashMask, UInt32 *heads, UInt32 numHeads, const UInt32 *crc); + +typedef struct _CMatchFinderMt +{ + /* LZ */ + const Byte *pointerToCurPos; + UInt32 *btBuf; + const UInt32 *btBufPos; + const UInt32 *btBufPosLimit; + UInt32 lzPos; + UInt32 btNumAvailBytes; + + UInt32 *hash; + UInt32 fixedHashSize; + // UInt32 hash4Mask; + UInt32 historySize; + const UInt32 *crc; + + Mf_Mix_Matches MixMatchesFunc; + UInt32 failure_LZ_BT; // failure in BT transfered to LZ + // UInt32 failure_LZ_LZ; // failure in LZ tables + UInt32 failureBuf[1]; + // UInt32 crc[256]; + + /* LZ + BT */ + CMtSync btSync; + Byte btDummy[kMtCacheLineDummy]; + + /* BT */ + UInt32 *hashBuf; + UInt32 hashBufPos; + UInt32 hashBufPosLimit; + UInt32 hashNumAvail; + UInt32 failure_BT; + + + CLzRef *son; + UInt32 matchMaxLen; + UInt32 numHashBytes; + UInt32 pos; + const Byte *buffer; + UInt32 cyclicBufferPos; + UInt32 cyclicBufferSize; /* it must be = (historySize + 1) */ + UInt32 cutValue; + + /* BT + Hash */ + CMtSync hashSync; + /* Byte hashDummy[kMtCacheLineDummy]; */ + + /* Hash */ + Mf_GetHeads GetHeadsFunc; + CMatchFinder *MatchFinder; + // CMatchFinder MatchFinder; +} CMatchFinderMt; + +// only for Mt part +void MatchFinderMt_Construct(CMatchFinderMt *p); +void MatchFinderMt_Destruct(CMatchFinderMt *p, ISzAllocPtr alloc); + +SRes MatchFinderMt_Create(CMatchFinderMt *p, UInt32 historySize, UInt32 keepAddBufferBefore, + UInt32 matchMaxLen, UInt32 keepAddBufferAfter, ISzAllocPtr alloc); +void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder2 *vTable); + +/* call MatchFinderMt_InitMt() before IMatchFinder::Init() */ +SRes MatchFinderMt_InitMt(CMatchFinderMt *p); +void MatchFinderMt_ReleaseStream(CMatchFinderMt *p); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/LzFindOpt.c b/deps/LZMA-SDK/C/LzFindOpt.c index dbb6ad9fc..8ff006e07 100644 --- a/deps/LZMA-SDK/C/LzFindOpt.c +++ b/deps/LZMA-SDK/C/LzFindOpt.c @@ -1,578 +1,578 @@ -/* LzFindOpt.c -- multithreaded Match finder for LZ algorithms -2021-07-13 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "CpuArch.h" -#include "LzFind.h" - -// #include "LzFindMt.h" - -// #define LOG_ITERS - -// #define LOG_THREAD - -#ifdef LOG_THREAD -#include -#define PRF(x) x -#else -// #define PRF(x) -#endif - -#ifdef LOG_ITERS -#include -UInt64 g_NumIters_Tree; -UInt64 g_NumIters_Loop; -UInt64 g_NumIters_Bytes; -#define LOG_ITER(x) x -#else -#define LOG_ITER(x) -#endif - -// ---------- BT THREAD ---------- - -#define USE_SON_PREFETCH -#define USE_LONG_MATCH_OPT - -#define kEmptyHashValue 0 - -// #define CYC_TO_POS_OFFSET 0 - -// #define CYC_TO_POS_OFFSET 1 // for debug - -/* -MY_NO_INLINE -UInt32 * MY_FAST_CALL GetMatchesSpecN_1(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, - UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, UInt32 *posRes) -{ - do - { - UInt32 delta; - if (hash == size) - break; - delta = *hash++; - - if (delta == 0 || delta > (UInt32)pos) - return NULL; - - lenLimit++; - - if (delta == (UInt32)pos) - { - CLzRef *ptr1 = son + ((size_t)pos << 1) - CYC_TO_POS_OFFSET * 2; - *d++ = 0; - ptr1[0] = kEmptyHashValue; - ptr1[1] = kEmptyHashValue; - } -else -{ - UInt32 *_distances = ++d; - - CLzRef *ptr0 = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2 + 1; - CLzRef *ptr1 = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2; - - const Byte *len0 = cur, *len1 = cur; - UInt32 cutValue = _cutValue; - const Byte *maxLen = cur + _maxLen; - - for (LOG_ITER(g_NumIters_Tree++);;) - { - LOG_ITER(g_NumIters_Loop++); - { - const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; - CLzRef *pair = son + ((size_t)(((ptrdiff_t)pos - CYC_TO_POS_OFFSET) + diff) << 1); - const Byte *len = (len0 < len1 ? len0 : len1); - - #ifdef USE_SON_PREFETCH - const UInt32 pair0 = *pair; - #endif - - if (len[diff] == len[0]) - { - if (++len != lenLimit && len[diff] == len[0]) - while (++len != lenLimit) - { - LOG_ITER(g_NumIters_Bytes++); - if (len[diff] != len[0]) - break; - } - if (maxLen < len) - { - maxLen = len; - *d++ = (UInt32)(len - cur); - *d++ = delta - 1; - - if (len == lenLimit) - { - const UInt32 pair1 = pair[1]; - *ptr1 = - #ifdef USE_SON_PREFETCH - pair0; - #else - pair[0]; - #endif - *ptr0 = pair1; - - _distances[-1] = (UInt32)(d - _distances); - - #ifdef USE_LONG_MATCH_OPT - - if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) - break; - - { - for (;;) - { - hash++; - pos++; - cur++; - lenLimit++; - { - CLzRef *ptr = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2; - #if 0 - *(UInt64 *)(void *)ptr = ((const UInt64 *)(const void *)ptr)[diff]; - #else - const UInt32 p0 = ptr[0 + (diff * 2)]; - const UInt32 p1 = ptr[1 + (diff * 2)]; - ptr[0] = p0; - ptr[1] = p1; - // ptr[0] = ptr[0 + (diff * 2)]; - // ptr[1] = ptr[1 + (diff * 2)]; - #endif - } - // PrintSon(son + 2, pos - 1); - // printf("\npos = %x delta = %x\n", pos, delta); - len++; - *d++ = 2; - *d++ = (UInt32)(len - cur); - *d++ = delta - 1; - if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) - break; - } - } - #endif - - break; - } - } - } - - { - const UInt32 curMatch = (UInt32)pos - delta; // (UInt32)(pos + diff); - if (len[diff] < len[0]) - { - delta = pair[1]; - if (delta >= curMatch) - return NULL; - *ptr1 = curMatch; - ptr1 = pair + 1; - len1 = len; - } - else - { - delta = *pair; - if (delta >= curMatch) - return NULL; - *ptr0 = curMatch; - ptr0 = pair; - len0 = len; - } - - delta = (UInt32)pos - delta; - - if (--cutValue == 0 || delta >= pos) - { - *ptr0 = *ptr1 = kEmptyHashValue; - _distances[-1] = (UInt32)(d - _distances); - break; - } - } - } - } // for (tree iterations) -} - pos++; - cur++; - } - while (d < limit); - *posRes = (UInt32)pos; - return d; -} -*/ - -/* define cbs if you use 2 functions. - GetMatchesSpecN_1() : (pos < _cyclicBufferSize) - GetMatchesSpecN_2() : (pos >= _cyclicBufferSize) - - do not define cbs if you use 1 function: - GetMatchesSpecN_2() -*/ - -// #define cbs _cyclicBufferSize - -/* - we use size_t for (pos) and (_cyclicBufferPos_ instead of UInt32 - to eliminate "movsx" BUG in old MSVC x64 compiler. -*/ - -UInt32 * MY_FAST_CALL GetMatchesSpecN_2(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, - UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, - size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, - UInt32 *posRes); - -MY_NO_INLINE -UInt32 * MY_FAST_CALL GetMatchesSpecN_2(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, - UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, - size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, - UInt32 *posRes) -{ - do // while (hash != size) - { - UInt32 delta; - - #ifndef cbs - UInt32 cbs; - #endif - - if (hash == size) - break; - - delta = *hash++; - - if (delta == 0) - return NULL; - - lenLimit++; - - #ifndef cbs - cbs = _cyclicBufferSize; - if ((UInt32)pos < cbs) - { - if (delta > (UInt32)pos) - return NULL; - cbs = (UInt32)pos; - } - #endif - - if (delta >= cbs) - { - CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); - *d++ = 0; - ptr1[0] = kEmptyHashValue; - ptr1[1] = kEmptyHashValue; - } -else -{ - UInt32 *_distances = ++d; - - CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; - CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); - - UInt32 cutValue = _cutValue; - const Byte *len0 = cur, *len1 = cur; - const Byte *maxLen = cur + _maxLen; - - // if (cutValue == 0) { *ptr0 = *ptr1 = kEmptyHashValue; } else - for (LOG_ITER(g_NumIters_Tree++);;) - { - LOG_ITER(g_NumIters_Loop++); - { - // SPEC code - CLzRef *pair = son + ((size_t)((ptrdiff_t)_cyclicBufferPos - (ptrdiff_t)delta - + (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0) - ) << 1); - - const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; - const Byte *len = (len0 < len1 ? len0 : len1); - - #ifdef USE_SON_PREFETCH - const UInt32 pair0 = *pair; - #endif - - if (len[diff] == len[0]) - { - if (++len != lenLimit && len[diff] == len[0]) - while (++len != lenLimit) - { - LOG_ITER(g_NumIters_Bytes++); - if (len[diff] != len[0]) - break; - } - if (maxLen < len) - { - maxLen = len; - *d++ = (UInt32)(len - cur); - *d++ = delta - 1; - - if (len == lenLimit) - { - const UInt32 pair1 = pair[1]; - *ptr1 = - #ifdef USE_SON_PREFETCH - pair0; - #else - pair[0]; - #endif - *ptr0 = pair1; - - _distances[-1] = (UInt32)(d - _distances); - - #ifdef USE_LONG_MATCH_OPT - - if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) - break; - - { - for (;;) - { - *d++ = 2; - *d++ = (UInt32)(lenLimit - cur); - *d++ = delta - 1; - cur++; - lenLimit++; - // SPEC - _cyclicBufferPos++; - { - // SPEC code - CLzRef *dest = son + ((size_t)(_cyclicBufferPos) << 1); - const CLzRef *src = dest + ((diff - + (ptrdiff_t)(UInt32)((_cyclicBufferPos < delta) ? cbs : 0)) << 1); - // CLzRef *ptr = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2; - #if 0 - *(UInt64 *)(void *)dest = *((const UInt64 *)(const void *)src); - #else - const UInt32 p0 = src[0]; - const UInt32 p1 = src[1]; - dest[0] = p0; - dest[1] = p1; - #endif - } - pos++; - hash++; - if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) - break; - } // for() end for long matches - } - #endif - - break; // break from TREE iterations - } - } - } - { - const UInt32 curMatch = (UInt32)pos - delta; // (UInt32)(pos + diff); - if (len[diff] < len[0]) - { - delta = pair[1]; - *ptr1 = curMatch; - ptr1 = pair + 1; - len1 = len; - if (delta >= curMatch) - return NULL; - } - else - { - delta = *pair; - *ptr0 = curMatch; - ptr0 = pair; - len0 = len; - if (delta >= curMatch) - return NULL; - } - delta = (UInt32)pos - delta; - - if (--cutValue == 0 || delta >= cbs) - { - *ptr0 = *ptr1 = kEmptyHashValue; - _distances[-1] = (UInt32)(d - _distances); - break; - } - } - } - } // for (tree iterations) -} - pos++; - _cyclicBufferPos++; - cur++; - } - while (d < limit); - *posRes = (UInt32)pos; - return d; -} - - - -/* -typedef UInt32 uint32plus; // size_t - -UInt32 * MY_FAST_CALL GetMatchesSpecN_3(uint32plus lenLimit, size_t pos, const Byte *cur, CLzRef *son, - UInt32 _cutValue, UInt32 *d, uint32plus _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, - size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, - UInt32 *posRes) -{ - do // while (hash != size) - { - UInt32 delta; - - #ifndef cbs - UInt32 cbs; - #endif - - if (hash == size) - break; - - delta = *hash++; - - if (delta == 0) - return NULL; - - #ifndef cbs - cbs = _cyclicBufferSize; - if ((UInt32)pos < cbs) - { - if (delta > (UInt32)pos) - return NULL; - cbs = (UInt32)pos; - } - #endif - - if (delta >= cbs) - { - CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); - *d++ = 0; - ptr1[0] = kEmptyHashValue; - ptr1[1] = kEmptyHashValue; - } -else -{ - CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; - CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); - UInt32 *_distances = ++d; - uint32plus len0 = 0, len1 = 0; - UInt32 cutValue = _cutValue; - uint32plus maxLen = _maxLen; - // lenLimit++; // const Byte *lenLimit = cur + _lenLimit; - - for (LOG_ITER(g_NumIters_Tree++);;) - { - LOG_ITER(g_NumIters_Loop++); - { - // const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; - CLzRef *pair = son + ((size_t)((ptrdiff_t)_cyclicBufferPos - delta - + (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0) - ) << 1); - const Byte *pb = cur - delta; - uint32plus len = (len0 < len1 ? len0 : len1); - - #ifdef USE_SON_PREFETCH - const UInt32 pair0 = *pair; - #endif - - if (pb[len] == cur[len]) - { - if (++len != lenLimit && pb[len] == cur[len]) - while (++len != lenLimit) - if (pb[len] != cur[len]) - break; - if (maxLen < len) - { - maxLen = len; - *d++ = (UInt32)len; - *d++ = delta - 1; - if (len == lenLimit) - { - { - const UInt32 pair1 = pair[1]; - *ptr0 = pair1; - *ptr1 = - #ifdef USE_SON_PREFETCH - pair0; - #else - pair[0]; - #endif - } - - _distances[-1] = (UInt32)(d - _distances); - - #ifdef USE_LONG_MATCH_OPT - - if (hash == size || *hash != delta || pb[lenLimit] != cur[lenLimit] || d >= limit) - break; - - { - const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; - for (;;) - { - *d++ = 2; - *d++ = (UInt32)lenLimit; - *d++ = delta - 1; - _cyclicBufferPos++; - { - CLzRef *dest = son + ((size_t)_cyclicBufferPos << 1); - const CLzRef *src = dest + ((diff + - (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0)) << 1); - #if 0 - *(UInt64 *)(void *)dest = *((const UInt64 *)(const void *)src); - #else - const UInt32 p0 = src[0]; - const UInt32 p1 = src[1]; - dest[0] = p0; - dest[1] = p1; - #endif - } - hash++; - pos++; - cur++; - pb++; - if (hash == size || *hash != delta || pb[lenLimit] != cur[lenLimit] || d >= limit) - break; - } - } - #endif - - break; - } - } - } - { - const UInt32 curMatch = (UInt32)pos - delta; - if (pb[len] < cur[len]) - { - delta = pair[1]; - *ptr1 = curMatch; - ptr1 = pair + 1; - len1 = len; - } - else - { - delta = *pair; - *ptr0 = curMatch; - ptr0 = pair; - len0 = len; - } - - { - if (delta >= curMatch) - return NULL; - delta = (UInt32)pos - delta; - if (delta >= cbs - // delta >= _cyclicBufferSize || delta >= pos - || --cutValue == 0) - { - *ptr0 = *ptr1 = kEmptyHashValue; - _distances[-1] = (UInt32)(d - _distances); - break; - } - } - } - } - } // for (tree iterations) -} - pos++; - _cyclicBufferPos++; - cur++; - } - while (d < limit); - *posRes = (UInt32)pos; - return d; -} -*/ +/* LzFindOpt.c -- multithreaded Match finder for LZ algorithms +2021-07-13 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" +#include "LzFind.h" + +// #include "LzFindMt.h" + +// #define LOG_ITERS + +// #define LOG_THREAD + +#ifdef LOG_THREAD +#include +#define PRF(x) x +#else +// #define PRF(x) +#endif + +#ifdef LOG_ITERS +#include +UInt64 g_NumIters_Tree; +UInt64 g_NumIters_Loop; +UInt64 g_NumIters_Bytes; +#define LOG_ITER(x) x +#else +#define LOG_ITER(x) +#endif + +// ---------- BT THREAD ---------- + +#define USE_SON_PREFETCH +#define USE_LONG_MATCH_OPT + +#define kEmptyHashValue 0 + +// #define CYC_TO_POS_OFFSET 0 + +// #define CYC_TO_POS_OFFSET 1 // for debug + +/* +MY_NO_INLINE +UInt32 * MY_FAST_CALL GetMatchesSpecN_1(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, + UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, UInt32 *posRes) +{ + do + { + UInt32 delta; + if (hash == size) + break; + delta = *hash++; + + if (delta == 0 || delta > (UInt32)pos) + return NULL; + + lenLimit++; + + if (delta == (UInt32)pos) + { + CLzRef *ptr1 = son + ((size_t)pos << 1) - CYC_TO_POS_OFFSET * 2; + *d++ = 0; + ptr1[0] = kEmptyHashValue; + ptr1[1] = kEmptyHashValue; + } +else +{ + UInt32 *_distances = ++d; + + CLzRef *ptr0 = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2 + 1; + CLzRef *ptr1 = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2; + + const Byte *len0 = cur, *len1 = cur; + UInt32 cutValue = _cutValue; + const Byte *maxLen = cur + _maxLen; + + for (LOG_ITER(g_NumIters_Tree++);;) + { + LOG_ITER(g_NumIters_Loop++); + { + const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; + CLzRef *pair = son + ((size_t)(((ptrdiff_t)pos - CYC_TO_POS_OFFSET) + diff) << 1); + const Byte *len = (len0 < len1 ? len0 : len1); + + #ifdef USE_SON_PREFETCH + const UInt32 pair0 = *pair; + #endif + + if (len[diff] == len[0]) + { + if (++len != lenLimit && len[diff] == len[0]) + while (++len != lenLimit) + { + LOG_ITER(g_NumIters_Bytes++); + if (len[diff] != len[0]) + break; + } + if (maxLen < len) + { + maxLen = len; + *d++ = (UInt32)(len - cur); + *d++ = delta - 1; + + if (len == lenLimit) + { + const UInt32 pair1 = pair[1]; + *ptr1 = + #ifdef USE_SON_PREFETCH + pair0; + #else + pair[0]; + #endif + *ptr0 = pair1; + + _distances[-1] = (UInt32)(d - _distances); + + #ifdef USE_LONG_MATCH_OPT + + if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) + break; + + { + for (;;) + { + hash++; + pos++; + cur++; + lenLimit++; + { + CLzRef *ptr = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2; + #if 0 + *(UInt64 *)(void *)ptr = ((const UInt64 *)(const void *)ptr)[diff]; + #else + const UInt32 p0 = ptr[0 + (diff * 2)]; + const UInt32 p1 = ptr[1 + (diff * 2)]; + ptr[0] = p0; + ptr[1] = p1; + // ptr[0] = ptr[0 + (diff * 2)]; + // ptr[1] = ptr[1 + (diff * 2)]; + #endif + } + // PrintSon(son + 2, pos - 1); + // printf("\npos = %x delta = %x\n", pos, delta); + len++; + *d++ = 2; + *d++ = (UInt32)(len - cur); + *d++ = delta - 1; + if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) + break; + } + } + #endif + + break; + } + } + } + + { + const UInt32 curMatch = (UInt32)pos - delta; // (UInt32)(pos + diff); + if (len[diff] < len[0]) + { + delta = pair[1]; + if (delta >= curMatch) + return NULL; + *ptr1 = curMatch; + ptr1 = pair + 1; + len1 = len; + } + else + { + delta = *pair; + if (delta >= curMatch) + return NULL; + *ptr0 = curMatch; + ptr0 = pair; + len0 = len; + } + + delta = (UInt32)pos - delta; + + if (--cutValue == 0 || delta >= pos) + { + *ptr0 = *ptr1 = kEmptyHashValue; + _distances[-1] = (UInt32)(d - _distances); + break; + } + } + } + } // for (tree iterations) +} + pos++; + cur++; + } + while (d < limit); + *posRes = (UInt32)pos; + return d; +} +*/ + +/* define cbs if you use 2 functions. + GetMatchesSpecN_1() : (pos < _cyclicBufferSize) + GetMatchesSpecN_2() : (pos >= _cyclicBufferSize) + + do not define cbs if you use 1 function: + GetMatchesSpecN_2() +*/ + +// #define cbs _cyclicBufferSize + +/* + we use size_t for (pos) and (_cyclicBufferPos_ instead of UInt32 + to eliminate "movsx" BUG in old MSVC x64 compiler. +*/ + +UInt32 * MY_FAST_CALL GetMatchesSpecN_2(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, + UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, + UInt32 *posRes); + +MY_NO_INLINE +UInt32 * MY_FAST_CALL GetMatchesSpecN_2(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, + UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, + UInt32 *posRes) +{ + do // while (hash != size) + { + UInt32 delta; + + #ifndef cbs + UInt32 cbs; + #endif + + if (hash == size) + break; + + delta = *hash++; + + if (delta == 0) + return NULL; + + lenLimit++; + + #ifndef cbs + cbs = _cyclicBufferSize; + if ((UInt32)pos < cbs) + { + if (delta > (UInt32)pos) + return NULL; + cbs = (UInt32)pos; + } + #endif + + if (delta >= cbs) + { + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + *d++ = 0; + ptr1[0] = kEmptyHashValue; + ptr1[1] = kEmptyHashValue; + } +else +{ + UInt32 *_distances = ++d; + + CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + + UInt32 cutValue = _cutValue; + const Byte *len0 = cur, *len1 = cur; + const Byte *maxLen = cur + _maxLen; + + // if (cutValue == 0) { *ptr0 = *ptr1 = kEmptyHashValue; } else + for (LOG_ITER(g_NumIters_Tree++);;) + { + LOG_ITER(g_NumIters_Loop++); + { + // SPEC code + CLzRef *pair = son + ((size_t)((ptrdiff_t)_cyclicBufferPos - (ptrdiff_t)delta + + (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0) + ) << 1); + + const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; + const Byte *len = (len0 < len1 ? len0 : len1); + + #ifdef USE_SON_PREFETCH + const UInt32 pair0 = *pair; + #endif + + if (len[diff] == len[0]) + { + if (++len != lenLimit && len[diff] == len[0]) + while (++len != lenLimit) + { + LOG_ITER(g_NumIters_Bytes++); + if (len[diff] != len[0]) + break; + } + if (maxLen < len) + { + maxLen = len; + *d++ = (UInt32)(len - cur); + *d++ = delta - 1; + + if (len == lenLimit) + { + const UInt32 pair1 = pair[1]; + *ptr1 = + #ifdef USE_SON_PREFETCH + pair0; + #else + pair[0]; + #endif + *ptr0 = pair1; + + _distances[-1] = (UInt32)(d - _distances); + + #ifdef USE_LONG_MATCH_OPT + + if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) + break; + + { + for (;;) + { + *d++ = 2; + *d++ = (UInt32)(lenLimit - cur); + *d++ = delta - 1; + cur++; + lenLimit++; + // SPEC + _cyclicBufferPos++; + { + // SPEC code + CLzRef *dest = son + ((size_t)(_cyclicBufferPos) << 1); + const CLzRef *src = dest + ((diff + + (ptrdiff_t)(UInt32)((_cyclicBufferPos < delta) ? cbs : 0)) << 1); + // CLzRef *ptr = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2; + #if 0 + *(UInt64 *)(void *)dest = *((const UInt64 *)(const void *)src); + #else + const UInt32 p0 = src[0]; + const UInt32 p1 = src[1]; + dest[0] = p0; + dest[1] = p1; + #endif + } + pos++; + hash++; + if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) + break; + } // for() end for long matches + } + #endif + + break; // break from TREE iterations + } + } + } + { + const UInt32 curMatch = (UInt32)pos - delta; // (UInt32)(pos + diff); + if (len[diff] < len[0]) + { + delta = pair[1]; + *ptr1 = curMatch; + ptr1 = pair + 1; + len1 = len; + if (delta >= curMatch) + return NULL; + } + else + { + delta = *pair; + *ptr0 = curMatch; + ptr0 = pair; + len0 = len; + if (delta >= curMatch) + return NULL; + } + delta = (UInt32)pos - delta; + + if (--cutValue == 0 || delta >= cbs) + { + *ptr0 = *ptr1 = kEmptyHashValue; + _distances[-1] = (UInt32)(d - _distances); + break; + } + } + } + } // for (tree iterations) +} + pos++; + _cyclicBufferPos++; + cur++; + } + while (d < limit); + *posRes = (UInt32)pos; + return d; +} + + + +/* +typedef UInt32 uint32plus; // size_t + +UInt32 * MY_FAST_CALL GetMatchesSpecN_3(uint32plus lenLimit, size_t pos, const Byte *cur, CLzRef *son, + UInt32 _cutValue, UInt32 *d, uint32plus _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, + UInt32 *posRes) +{ + do // while (hash != size) + { + UInt32 delta; + + #ifndef cbs + UInt32 cbs; + #endif + + if (hash == size) + break; + + delta = *hash++; + + if (delta == 0) + return NULL; + + #ifndef cbs + cbs = _cyclicBufferSize; + if ((UInt32)pos < cbs) + { + if (delta > (UInt32)pos) + return NULL; + cbs = (UInt32)pos; + } + #endif + + if (delta >= cbs) + { + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + *d++ = 0; + ptr1[0] = kEmptyHashValue; + ptr1[1] = kEmptyHashValue; + } +else +{ + CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + UInt32 *_distances = ++d; + uint32plus len0 = 0, len1 = 0; + UInt32 cutValue = _cutValue; + uint32plus maxLen = _maxLen; + // lenLimit++; // const Byte *lenLimit = cur + _lenLimit; + + for (LOG_ITER(g_NumIters_Tree++);;) + { + LOG_ITER(g_NumIters_Loop++); + { + // const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; + CLzRef *pair = son + ((size_t)((ptrdiff_t)_cyclicBufferPos - delta + + (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0) + ) << 1); + const Byte *pb = cur - delta; + uint32plus len = (len0 < len1 ? len0 : len1); + + #ifdef USE_SON_PREFETCH + const UInt32 pair0 = *pair; + #endif + + if (pb[len] == cur[len]) + { + if (++len != lenLimit && pb[len] == cur[len]) + while (++len != lenLimit) + if (pb[len] != cur[len]) + break; + if (maxLen < len) + { + maxLen = len; + *d++ = (UInt32)len; + *d++ = delta - 1; + if (len == lenLimit) + { + { + const UInt32 pair1 = pair[1]; + *ptr0 = pair1; + *ptr1 = + #ifdef USE_SON_PREFETCH + pair0; + #else + pair[0]; + #endif + } + + _distances[-1] = (UInt32)(d - _distances); + + #ifdef USE_LONG_MATCH_OPT + + if (hash == size || *hash != delta || pb[lenLimit] != cur[lenLimit] || d >= limit) + break; + + { + const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; + for (;;) + { + *d++ = 2; + *d++ = (UInt32)lenLimit; + *d++ = delta - 1; + _cyclicBufferPos++; + { + CLzRef *dest = son + ((size_t)_cyclicBufferPos << 1); + const CLzRef *src = dest + ((diff + + (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0)) << 1); + #if 0 + *(UInt64 *)(void *)dest = *((const UInt64 *)(const void *)src); + #else + const UInt32 p0 = src[0]; + const UInt32 p1 = src[1]; + dest[0] = p0; + dest[1] = p1; + #endif + } + hash++; + pos++; + cur++; + pb++; + if (hash == size || *hash != delta || pb[lenLimit] != cur[lenLimit] || d >= limit) + break; + } + } + #endif + + break; + } + } + } + { + const UInt32 curMatch = (UInt32)pos - delta; + if (pb[len] < cur[len]) + { + delta = pair[1]; + *ptr1 = curMatch; + ptr1 = pair + 1; + len1 = len; + } + else + { + delta = *pair; + *ptr0 = curMatch; + ptr0 = pair; + len0 = len; + } + + { + if (delta >= curMatch) + return NULL; + delta = (UInt32)pos - delta; + if (delta >= cbs + // delta >= _cyclicBufferSize || delta >= pos + || --cutValue == 0) + { + *ptr0 = *ptr1 = kEmptyHashValue; + _distances[-1] = (UInt32)(d - _distances); + break; + } + } + } + } + } // for (tree iterations) +} + pos++; + _cyclicBufferPos++; + cur++; + } + while (d < limit); + *posRes = (UInt32)pos; + return d; +} +*/ diff --git a/deps/LZMA-SDK/C/LzHash.h b/deps/LZMA-SDK/C/LzHash.h index a682f83be..77b898cfa 100644 --- a/deps/LZMA-SDK/C/LzHash.h +++ b/deps/LZMA-SDK/C/LzHash.h @@ -1,34 +1,34 @@ -/* LzHash.h -- HASH functions for LZ algorithms -2019-10-30 : Igor Pavlov : Public domain */ - -#ifndef __LZ_HASH_H -#define __LZ_HASH_H - -/* - (kHash2Size >= (1 << 8)) : Required - (kHash3Size >= (1 << 16)) : Required -*/ - -#define kHash2Size (1 << 10) -#define kHash3Size (1 << 16) -// #define kHash4Size (1 << 20) - -#define kFix3HashSize (kHash2Size) -#define kFix4HashSize (kHash2Size + kHash3Size) -// #define kFix5HashSize (kHash2Size + kHash3Size + kHash4Size) - -/* - We use up to 3 crc values for hash: - crc0 - crc1 << Shift_1 - crc2 << Shift_2 - (Shift_1 = 5) and (Shift_2 = 10) is good tradeoff. - Small values for Shift are not good for collision rate. - Big value for Shift_2 increases the minimum size - of hash table, that will be slow for small files. -*/ - -#define kLzHash_CrcShift_1 5 -#define kLzHash_CrcShift_2 10 - -#endif +/* LzHash.h -- HASH functions for LZ algorithms +2019-10-30 : Igor Pavlov : Public domain */ + +#ifndef __LZ_HASH_H +#define __LZ_HASH_H + +/* + (kHash2Size >= (1 << 8)) : Required + (kHash3Size >= (1 << 16)) : Required +*/ + +#define kHash2Size (1 << 10) +#define kHash3Size (1 << 16) +// #define kHash4Size (1 << 20) + +#define kFix3HashSize (kHash2Size) +#define kFix4HashSize (kHash2Size + kHash3Size) +// #define kFix5HashSize (kHash2Size + kHash3Size + kHash4Size) + +/* + We use up to 3 crc values for hash: + crc0 + crc1 << Shift_1 + crc2 << Shift_2 + (Shift_1 = 5) and (Shift_2 = 10) is good tradeoff. + Small values for Shift are not good for collision rate. + Big value for Shift_2 increases the minimum size + of hash table, that will be slow for small files. +*/ + +#define kLzHash_CrcShift_1 5 +#define kLzHash_CrcShift_2 10 + +#endif diff --git a/deps/LZMA-SDK/C/Lzma2Dec.c b/deps/LZMA-SDK/C/Lzma2Dec.c index f9f98095d..ac970a843 100644 --- a/deps/LZMA-SDK/C/Lzma2Dec.c +++ b/deps/LZMA-SDK/C/Lzma2Dec.c @@ -1,489 +1,489 @@ -/* Lzma2Dec.c -- LZMA2 Decoder -2021-02-09 : Igor Pavlov : Public domain */ - -/* #define SHOW_DEBUG_INFO */ - -#include "Precomp.h" - -#ifdef SHOW_DEBUG_INFO -#include -#endif - -#include - -#include "Lzma2Dec.h" - -/* -00000000 - End of data -00000001 U U - Uncompressed, reset dic, need reset state and set new prop -00000010 U U - Uncompressed, no reset -100uuuuu U U P P - LZMA, no reset -101uuuuu U U P P - LZMA, reset state -110uuuuu U U P P S - LZMA, reset state + set new prop -111uuuuu U U P P S - LZMA, reset state + set new prop, reset dic - - u, U - Unpack Size - P - Pack Size - S - Props -*/ - -#define LZMA2_CONTROL_COPY_RESET_DIC 1 - -#define LZMA2_IS_UNCOMPRESSED_STATE(p) (((p)->control & (1 << 7)) == 0) - -#define LZMA2_LCLP_MAX 4 -#define LZMA2_DIC_SIZE_FROM_PROP(p) (((UInt32)2 | ((p) & 1)) << ((p) / 2 + 11)) - -#ifdef SHOW_DEBUG_INFO -#define PRF(x) x -#else -#define PRF(x) -#endif - -typedef enum -{ - LZMA2_STATE_CONTROL, - LZMA2_STATE_UNPACK0, - LZMA2_STATE_UNPACK1, - LZMA2_STATE_PACK0, - LZMA2_STATE_PACK1, - LZMA2_STATE_PROP, - LZMA2_STATE_DATA, - LZMA2_STATE_DATA_CONT, - LZMA2_STATE_FINISHED, - LZMA2_STATE_ERROR -} ELzma2State; - -static SRes Lzma2Dec_GetOldProps(Byte prop, Byte *props) -{ - UInt32 dicSize; - if (prop > 40) - return SZ_ERROR_UNSUPPORTED; - dicSize = (prop == 40) ? 0xFFFFFFFF : LZMA2_DIC_SIZE_FROM_PROP(prop); - props[0] = (Byte)LZMA2_LCLP_MAX; - props[1] = (Byte)(dicSize); - props[2] = (Byte)(dicSize >> 8); - props[3] = (Byte)(dicSize >> 16); - props[4] = (Byte)(dicSize >> 24); - return SZ_OK; -} - -SRes Lzma2Dec_AllocateProbs(CLzma2Dec *p, Byte prop, ISzAllocPtr alloc) -{ - Byte props[LZMA_PROPS_SIZE]; - RINOK(Lzma2Dec_GetOldProps(prop, props)); - return LzmaDec_AllocateProbs(&p->decoder, props, LZMA_PROPS_SIZE, alloc); -} - -SRes Lzma2Dec_Allocate(CLzma2Dec *p, Byte prop, ISzAllocPtr alloc) -{ - Byte props[LZMA_PROPS_SIZE]; - RINOK(Lzma2Dec_GetOldProps(prop, props)); - return LzmaDec_Allocate(&p->decoder, props, LZMA_PROPS_SIZE, alloc); -} - -void Lzma2Dec_Init(CLzma2Dec *p) -{ - p->state = LZMA2_STATE_CONTROL; - p->needInitLevel = 0xE0; - p->isExtraMode = False; - p->unpackSize = 0; - - // p->decoder.dicPos = 0; // we can use it instead of full init - LzmaDec_Init(&p->decoder); -} - -// ELzma2State -static unsigned Lzma2Dec_UpdateState(CLzma2Dec *p, Byte b) -{ - switch (p->state) - { - case LZMA2_STATE_CONTROL: - p->isExtraMode = False; - p->control = b; - PRF(printf("\n %8X", (unsigned)p->decoder.dicPos)); - PRF(printf(" %02X", (unsigned)b)); - if (b == 0) - return LZMA2_STATE_FINISHED; - if (LZMA2_IS_UNCOMPRESSED_STATE(p)) - { - if (b == LZMA2_CONTROL_COPY_RESET_DIC) - p->needInitLevel = 0xC0; - else if (b > 2 || p->needInitLevel == 0xE0) - return LZMA2_STATE_ERROR; - } - else - { - if (b < p->needInitLevel) - return LZMA2_STATE_ERROR; - p->needInitLevel = 0; - p->unpackSize = (UInt32)(b & 0x1F) << 16; - } - return LZMA2_STATE_UNPACK0; - - case LZMA2_STATE_UNPACK0: - p->unpackSize |= (UInt32)b << 8; - return LZMA2_STATE_UNPACK1; - - case LZMA2_STATE_UNPACK1: - p->unpackSize |= (UInt32)b; - p->unpackSize++; - PRF(printf(" %7u", (unsigned)p->unpackSize)); - return LZMA2_IS_UNCOMPRESSED_STATE(p) ? LZMA2_STATE_DATA : LZMA2_STATE_PACK0; - - case LZMA2_STATE_PACK0: - p->packSize = (UInt32)b << 8; - return LZMA2_STATE_PACK1; - - case LZMA2_STATE_PACK1: - p->packSize |= (UInt32)b; - p->packSize++; - // if (p->packSize < 5) return LZMA2_STATE_ERROR; - PRF(printf(" %5u", (unsigned)p->packSize)); - return (p->control & 0x40) ? LZMA2_STATE_PROP : LZMA2_STATE_DATA; - - case LZMA2_STATE_PROP: - { - unsigned lc, lp; - if (b >= (9 * 5 * 5)) - return LZMA2_STATE_ERROR; - lc = b % 9; - b /= 9; - p->decoder.prop.pb = (Byte)(b / 5); - lp = b % 5; - if (lc + lp > LZMA2_LCLP_MAX) - return LZMA2_STATE_ERROR; - p->decoder.prop.lc = (Byte)lc; - p->decoder.prop.lp = (Byte)lp; - return LZMA2_STATE_DATA; - } - } - return LZMA2_STATE_ERROR; -} - -static void LzmaDec_UpdateWithUncompressed(CLzmaDec *p, const Byte *src, SizeT size) -{ - memcpy(p->dic + p->dicPos, src, size); - p->dicPos += size; - if (p->checkDicSize == 0 && p->prop.dicSize - p->processedPos <= size) - p->checkDicSize = p->prop.dicSize; - p->processedPos += (UInt32)size; -} - -void LzmaDec_InitDicAndState(CLzmaDec *p, BoolInt initDic, BoolInt initState); - - -SRes Lzma2Dec_DecodeToDic(CLzma2Dec *p, SizeT dicLimit, - const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status) -{ - SizeT inSize = *srcLen; - *srcLen = 0; - *status = LZMA_STATUS_NOT_SPECIFIED; - - while (p->state != LZMA2_STATE_ERROR) - { - SizeT dicPos; - - if (p->state == LZMA2_STATE_FINISHED) - { - *status = LZMA_STATUS_FINISHED_WITH_MARK; - return SZ_OK; - } - - dicPos = p->decoder.dicPos; - - if (dicPos == dicLimit && finishMode == LZMA_FINISH_ANY) - { - *status = LZMA_STATUS_NOT_FINISHED; - return SZ_OK; - } - - if (p->state != LZMA2_STATE_DATA && p->state != LZMA2_STATE_DATA_CONT) - { - if (*srcLen == inSize) - { - *status = LZMA_STATUS_NEEDS_MORE_INPUT; - return SZ_OK; - } - (*srcLen)++; - p->state = Lzma2Dec_UpdateState(p, *src++); - if (dicPos == dicLimit && p->state != LZMA2_STATE_FINISHED) - break; - continue; - } - - { - SizeT inCur = inSize - *srcLen; - SizeT outCur = dicLimit - dicPos; - ELzmaFinishMode curFinishMode = LZMA_FINISH_ANY; - - if (outCur >= p->unpackSize) - { - outCur = (SizeT)p->unpackSize; - curFinishMode = LZMA_FINISH_END; - } - - if (LZMA2_IS_UNCOMPRESSED_STATE(p)) - { - if (inCur == 0) - { - *status = LZMA_STATUS_NEEDS_MORE_INPUT; - return SZ_OK; - } - - if (p->state == LZMA2_STATE_DATA) - { - BoolInt initDic = (p->control == LZMA2_CONTROL_COPY_RESET_DIC); - LzmaDec_InitDicAndState(&p->decoder, initDic, False); - } - - if (inCur > outCur) - inCur = outCur; - if (inCur == 0) - break; - - LzmaDec_UpdateWithUncompressed(&p->decoder, src, inCur); - - src += inCur; - *srcLen += inCur; - p->unpackSize -= (UInt32)inCur; - p->state = (p->unpackSize == 0) ? LZMA2_STATE_CONTROL : LZMA2_STATE_DATA_CONT; - } - else - { - SRes res; - - if (p->state == LZMA2_STATE_DATA) - { - BoolInt initDic = (p->control >= 0xE0); - BoolInt initState = (p->control >= 0xA0); - LzmaDec_InitDicAndState(&p->decoder, initDic, initState); - p->state = LZMA2_STATE_DATA_CONT; - } - - if (inCur > p->packSize) - inCur = (SizeT)p->packSize; - - res = LzmaDec_DecodeToDic(&p->decoder, dicPos + outCur, src, &inCur, curFinishMode, status); - - src += inCur; - *srcLen += inCur; - p->packSize -= (UInt32)inCur; - outCur = p->decoder.dicPos - dicPos; - p->unpackSize -= (UInt32)outCur; - - if (res != 0) - break; - - if (*status == LZMA_STATUS_NEEDS_MORE_INPUT) - { - if (p->packSize == 0) - break; - return SZ_OK; - } - - if (inCur == 0 && outCur == 0) - { - if (*status != LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK - || p->unpackSize != 0 - || p->packSize != 0) - break; - p->state = LZMA2_STATE_CONTROL; - } - - *status = LZMA_STATUS_NOT_SPECIFIED; - } - } - } - - *status = LZMA_STATUS_NOT_SPECIFIED; - p->state = LZMA2_STATE_ERROR; - return SZ_ERROR_DATA; -} - - - - -ELzma2ParseStatus Lzma2Dec_Parse(CLzma2Dec *p, - SizeT outSize, - const Byte *src, SizeT *srcLen, - int checkFinishBlock) -{ - SizeT inSize = *srcLen; - *srcLen = 0; - - while (p->state != LZMA2_STATE_ERROR) - { - if (p->state == LZMA2_STATE_FINISHED) - return (ELzma2ParseStatus)LZMA_STATUS_FINISHED_WITH_MARK; - - if (outSize == 0 && !checkFinishBlock) - return (ELzma2ParseStatus)LZMA_STATUS_NOT_FINISHED; - - if (p->state != LZMA2_STATE_DATA && p->state != LZMA2_STATE_DATA_CONT) - { - if (*srcLen == inSize) - return (ELzma2ParseStatus)LZMA_STATUS_NEEDS_MORE_INPUT; - (*srcLen)++; - - p->state = Lzma2Dec_UpdateState(p, *src++); - - if (p->state == LZMA2_STATE_UNPACK0) - { - // if (p->decoder.dicPos != 0) - if (p->control == LZMA2_CONTROL_COPY_RESET_DIC || p->control >= 0xE0) - return LZMA2_PARSE_STATUS_NEW_BLOCK; - // if (outSize == 0) return LZMA_STATUS_NOT_FINISHED; - } - - // The following code can be commented. - // It's not big problem, if we read additional input bytes. - // It will be stopped later in LZMA2_STATE_DATA / LZMA2_STATE_DATA_CONT state. - - if (outSize == 0 && p->state != LZMA2_STATE_FINISHED) - { - // checkFinishBlock is true. So we expect that block must be finished, - // We can return LZMA_STATUS_NOT_SPECIFIED or LZMA_STATUS_NOT_FINISHED here - // break; - return (ELzma2ParseStatus)LZMA_STATUS_NOT_FINISHED; - } - - if (p->state == LZMA2_STATE_DATA) - return LZMA2_PARSE_STATUS_NEW_CHUNK; - - continue; - } - - if (outSize == 0) - return (ELzma2ParseStatus)LZMA_STATUS_NOT_FINISHED; - - { - SizeT inCur = inSize - *srcLen; - - if (LZMA2_IS_UNCOMPRESSED_STATE(p)) - { - if (inCur == 0) - return (ELzma2ParseStatus)LZMA_STATUS_NEEDS_MORE_INPUT; - if (inCur > p->unpackSize) - inCur = p->unpackSize; - if (inCur > outSize) - inCur = outSize; - p->decoder.dicPos += inCur; - src += inCur; - *srcLen += inCur; - outSize -= inCur; - p->unpackSize -= (UInt32)inCur; - p->state = (p->unpackSize == 0) ? LZMA2_STATE_CONTROL : LZMA2_STATE_DATA_CONT; - } - else - { - p->isExtraMode = True; - - if (inCur == 0) - { - if (p->packSize != 0) - return (ELzma2ParseStatus)LZMA_STATUS_NEEDS_MORE_INPUT; - } - else if (p->state == LZMA2_STATE_DATA) - { - p->state = LZMA2_STATE_DATA_CONT; - if (*src != 0) - { - // first byte of lzma chunk must be Zero - *srcLen += 1; - p->packSize--; - break; - } - } - - if (inCur > p->packSize) - inCur = (SizeT)p->packSize; - - src += inCur; - *srcLen += inCur; - p->packSize -= (UInt32)inCur; - - if (p->packSize == 0) - { - SizeT rem = outSize; - if (rem > p->unpackSize) - rem = p->unpackSize; - p->decoder.dicPos += rem; - p->unpackSize -= (UInt32)rem; - outSize -= rem; - if (p->unpackSize == 0) - p->state = LZMA2_STATE_CONTROL; - } - } - } - } - - p->state = LZMA2_STATE_ERROR; - return (ELzma2ParseStatus)LZMA_STATUS_NOT_SPECIFIED; -} - - - - -SRes Lzma2Dec_DecodeToBuf(CLzma2Dec *p, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status) -{ - SizeT outSize = *destLen, inSize = *srcLen; - *srcLen = *destLen = 0; - - for (;;) - { - SizeT inCur = inSize, outCur, dicPos; - ELzmaFinishMode curFinishMode; - SRes res; - - if (p->decoder.dicPos == p->decoder.dicBufSize) - p->decoder.dicPos = 0; - dicPos = p->decoder.dicPos; - curFinishMode = LZMA_FINISH_ANY; - outCur = p->decoder.dicBufSize - dicPos; - - if (outCur >= outSize) - { - outCur = outSize; - curFinishMode = finishMode; - } - - res = Lzma2Dec_DecodeToDic(p, dicPos + outCur, src, &inCur, curFinishMode, status); - - src += inCur; - inSize -= inCur; - *srcLen += inCur; - outCur = p->decoder.dicPos - dicPos; - memcpy(dest, p->decoder.dic + dicPos, outCur); - dest += outCur; - outSize -= outCur; - *destLen += outCur; - if (res != 0) - return res; - if (outCur == 0 || outSize == 0) - return SZ_OK; - } -} - - -SRes Lzma2Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, - Byte prop, ELzmaFinishMode finishMode, ELzmaStatus *status, ISzAllocPtr alloc) -{ - CLzma2Dec p; - SRes res; - SizeT outSize = *destLen, inSize = *srcLen; - *destLen = *srcLen = 0; - *status = LZMA_STATUS_NOT_SPECIFIED; - Lzma2Dec_Construct(&p); - RINOK(Lzma2Dec_AllocateProbs(&p, prop, alloc)); - p.decoder.dic = dest; - p.decoder.dicBufSize = outSize; - Lzma2Dec_Init(&p); - *srcLen = inSize; - res = Lzma2Dec_DecodeToDic(&p, outSize, src, srcLen, finishMode, status); - *destLen = p.decoder.dicPos; - if (res == SZ_OK && *status == LZMA_STATUS_NEEDS_MORE_INPUT) - res = SZ_ERROR_INPUT_EOF; - Lzma2Dec_FreeProbs(&p, alloc); - return res; -} +/* Lzma2Dec.c -- LZMA2 Decoder +2021-02-09 : Igor Pavlov : Public domain */ + +/* #define SHOW_DEBUG_INFO */ + +#include "Precomp.h" + +#ifdef SHOW_DEBUG_INFO +#include +#endif + +#include + +#include "Lzma2Dec.h" + +/* +00000000 - End of data +00000001 U U - Uncompressed, reset dic, need reset state and set new prop +00000010 U U - Uncompressed, no reset +100uuuuu U U P P - LZMA, no reset +101uuuuu U U P P - LZMA, reset state +110uuuuu U U P P S - LZMA, reset state + set new prop +111uuuuu U U P P S - LZMA, reset state + set new prop, reset dic + + u, U - Unpack Size + P - Pack Size + S - Props +*/ + +#define LZMA2_CONTROL_COPY_RESET_DIC 1 + +#define LZMA2_IS_UNCOMPRESSED_STATE(p) (((p)->control & (1 << 7)) == 0) + +#define LZMA2_LCLP_MAX 4 +#define LZMA2_DIC_SIZE_FROM_PROP(p) (((UInt32)2 | ((p) & 1)) << ((p) / 2 + 11)) + +#ifdef SHOW_DEBUG_INFO +#define PRF(x) x +#else +#define PRF(x) +#endif + +typedef enum +{ + LZMA2_STATE_CONTROL, + LZMA2_STATE_UNPACK0, + LZMA2_STATE_UNPACK1, + LZMA2_STATE_PACK0, + LZMA2_STATE_PACK1, + LZMA2_STATE_PROP, + LZMA2_STATE_DATA, + LZMA2_STATE_DATA_CONT, + LZMA2_STATE_FINISHED, + LZMA2_STATE_ERROR +} ELzma2State; + +static SRes Lzma2Dec_GetOldProps(Byte prop, Byte *props) +{ + UInt32 dicSize; + if (prop > 40) + return SZ_ERROR_UNSUPPORTED; + dicSize = (prop == 40) ? 0xFFFFFFFF : LZMA2_DIC_SIZE_FROM_PROP(prop); + props[0] = (Byte)LZMA2_LCLP_MAX; + props[1] = (Byte)(dicSize); + props[2] = (Byte)(dicSize >> 8); + props[3] = (Byte)(dicSize >> 16); + props[4] = (Byte)(dicSize >> 24); + return SZ_OK; +} + +SRes Lzma2Dec_AllocateProbs(CLzma2Dec *p, Byte prop, ISzAllocPtr alloc) +{ + Byte props[LZMA_PROPS_SIZE]; + RINOK(Lzma2Dec_GetOldProps(prop, props)); + return LzmaDec_AllocateProbs(&p->decoder, props, LZMA_PROPS_SIZE, alloc); +} + +SRes Lzma2Dec_Allocate(CLzma2Dec *p, Byte prop, ISzAllocPtr alloc) +{ + Byte props[LZMA_PROPS_SIZE]; + RINOK(Lzma2Dec_GetOldProps(prop, props)); + return LzmaDec_Allocate(&p->decoder, props, LZMA_PROPS_SIZE, alloc); +} + +void Lzma2Dec_Init(CLzma2Dec *p) +{ + p->state = LZMA2_STATE_CONTROL; + p->needInitLevel = 0xE0; + p->isExtraMode = False; + p->unpackSize = 0; + + // p->decoder.dicPos = 0; // we can use it instead of full init + LzmaDec_Init(&p->decoder); +} + +// ELzma2State +static unsigned Lzma2Dec_UpdateState(CLzma2Dec *p, Byte b) +{ + switch (p->state) + { + case LZMA2_STATE_CONTROL: + p->isExtraMode = False; + p->control = b; + PRF(printf("\n %8X", (unsigned)p->decoder.dicPos)); + PRF(printf(" %02X", (unsigned)b)); + if (b == 0) + return LZMA2_STATE_FINISHED; + if (LZMA2_IS_UNCOMPRESSED_STATE(p)) + { + if (b == LZMA2_CONTROL_COPY_RESET_DIC) + p->needInitLevel = 0xC0; + else if (b > 2 || p->needInitLevel == 0xE0) + return LZMA2_STATE_ERROR; + } + else + { + if (b < p->needInitLevel) + return LZMA2_STATE_ERROR; + p->needInitLevel = 0; + p->unpackSize = (UInt32)(b & 0x1F) << 16; + } + return LZMA2_STATE_UNPACK0; + + case LZMA2_STATE_UNPACK0: + p->unpackSize |= (UInt32)b << 8; + return LZMA2_STATE_UNPACK1; + + case LZMA2_STATE_UNPACK1: + p->unpackSize |= (UInt32)b; + p->unpackSize++; + PRF(printf(" %7u", (unsigned)p->unpackSize)); + return LZMA2_IS_UNCOMPRESSED_STATE(p) ? LZMA2_STATE_DATA : LZMA2_STATE_PACK0; + + case LZMA2_STATE_PACK0: + p->packSize = (UInt32)b << 8; + return LZMA2_STATE_PACK1; + + case LZMA2_STATE_PACK1: + p->packSize |= (UInt32)b; + p->packSize++; + // if (p->packSize < 5) return LZMA2_STATE_ERROR; + PRF(printf(" %5u", (unsigned)p->packSize)); + return (p->control & 0x40) ? LZMA2_STATE_PROP : LZMA2_STATE_DATA; + + case LZMA2_STATE_PROP: + { + unsigned lc, lp; + if (b >= (9 * 5 * 5)) + return LZMA2_STATE_ERROR; + lc = b % 9; + b /= 9; + p->decoder.prop.pb = (Byte)(b / 5); + lp = b % 5; + if (lc + lp > LZMA2_LCLP_MAX) + return LZMA2_STATE_ERROR; + p->decoder.prop.lc = (Byte)lc; + p->decoder.prop.lp = (Byte)lp; + return LZMA2_STATE_DATA; + } + } + return LZMA2_STATE_ERROR; +} + +static void LzmaDec_UpdateWithUncompressed(CLzmaDec *p, const Byte *src, SizeT size) +{ + memcpy(p->dic + p->dicPos, src, size); + p->dicPos += size; + if (p->checkDicSize == 0 && p->prop.dicSize - p->processedPos <= size) + p->checkDicSize = p->prop.dicSize; + p->processedPos += (UInt32)size; +} + +void LzmaDec_InitDicAndState(CLzmaDec *p, BoolInt initDic, BoolInt initState); + + +SRes Lzma2Dec_DecodeToDic(CLzma2Dec *p, SizeT dicLimit, + const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status) +{ + SizeT inSize = *srcLen; + *srcLen = 0; + *status = LZMA_STATUS_NOT_SPECIFIED; + + while (p->state != LZMA2_STATE_ERROR) + { + SizeT dicPos; + + if (p->state == LZMA2_STATE_FINISHED) + { + *status = LZMA_STATUS_FINISHED_WITH_MARK; + return SZ_OK; + } + + dicPos = p->decoder.dicPos; + + if (dicPos == dicLimit && finishMode == LZMA_FINISH_ANY) + { + *status = LZMA_STATUS_NOT_FINISHED; + return SZ_OK; + } + + if (p->state != LZMA2_STATE_DATA && p->state != LZMA2_STATE_DATA_CONT) + { + if (*srcLen == inSize) + { + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + (*srcLen)++; + p->state = Lzma2Dec_UpdateState(p, *src++); + if (dicPos == dicLimit && p->state != LZMA2_STATE_FINISHED) + break; + continue; + } + + { + SizeT inCur = inSize - *srcLen; + SizeT outCur = dicLimit - dicPos; + ELzmaFinishMode curFinishMode = LZMA_FINISH_ANY; + + if (outCur >= p->unpackSize) + { + outCur = (SizeT)p->unpackSize; + curFinishMode = LZMA_FINISH_END; + } + + if (LZMA2_IS_UNCOMPRESSED_STATE(p)) + { + if (inCur == 0) + { + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + + if (p->state == LZMA2_STATE_DATA) + { + BoolInt initDic = (p->control == LZMA2_CONTROL_COPY_RESET_DIC); + LzmaDec_InitDicAndState(&p->decoder, initDic, False); + } + + if (inCur > outCur) + inCur = outCur; + if (inCur == 0) + break; + + LzmaDec_UpdateWithUncompressed(&p->decoder, src, inCur); + + src += inCur; + *srcLen += inCur; + p->unpackSize -= (UInt32)inCur; + p->state = (p->unpackSize == 0) ? LZMA2_STATE_CONTROL : LZMA2_STATE_DATA_CONT; + } + else + { + SRes res; + + if (p->state == LZMA2_STATE_DATA) + { + BoolInt initDic = (p->control >= 0xE0); + BoolInt initState = (p->control >= 0xA0); + LzmaDec_InitDicAndState(&p->decoder, initDic, initState); + p->state = LZMA2_STATE_DATA_CONT; + } + + if (inCur > p->packSize) + inCur = (SizeT)p->packSize; + + res = LzmaDec_DecodeToDic(&p->decoder, dicPos + outCur, src, &inCur, curFinishMode, status); + + src += inCur; + *srcLen += inCur; + p->packSize -= (UInt32)inCur; + outCur = p->decoder.dicPos - dicPos; + p->unpackSize -= (UInt32)outCur; + + if (res != 0) + break; + + if (*status == LZMA_STATUS_NEEDS_MORE_INPUT) + { + if (p->packSize == 0) + break; + return SZ_OK; + } + + if (inCur == 0 && outCur == 0) + { + if (*status != LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK + || p->unpackSize != 0 + || p->packSize != 0) + break; + p->state = LZMA2_STATE_CONTROL; + } + + *status = LZMA_STATUS_NOT_SPECIFIED; + } + } + } + + *status = LZMA_STATUS_NOT_SPECIFIED; + p->state = LZMA2_STATE_ERROR; + return SZ_ERROR_DATA; +} + + + + +ELzma2ParseStatus Lzma2Dec_Parse(CLzma2Dec *p, + SizeT outSize, + const Byte *src, SizeT *srcLen, + int checkFinishBlock) +{ + SizeT inSize = *srcLen; + *srcLen = 0; + + while (p->state != LZMA2_STATE_ERROR) + { + if (p->state == LZMA2_STATE_FINISHED) + return (ELzma2ParseStatus)LZMA_STATUS_FINISHED_WITH_MARK; + + if (outSize == 0 && !checkFinishBlock) + return (ELzma2ParseStatus)LZMA_STATUS_NOT_FINISHED; + + if (p->state != LZMA2_STATE_DATA && p->state != LZMA2_STATE_DATA_CONT) + { + if (*srcLen == inSize) + return (ELzma2ParseStatus)LZMA_STATUS_NEEDS_MORE_INPUT; + (*srcLen)++; + + p->state = Lzma2Dec_UpdateState(p, *src++); + + if (p->state == LZMA2_STATE_UNPACK0) + { + // if (p->decoder.dicPos != 0) + if (p->control == LZMA2_CONTROL_COPY_RESET_DIC || p->control >= 0xE0) + return LZMA2_PARSE_STATUS_NEW_BLOCK; + // if (outSize == 0) return LZMA_STATUS_NOT_FINISHED; + } + + // The following code can be commented. + // It's not big problem, if we read additional input bytes. + // It will be stopped later in LZMA2_STATE_DATA / LZMA2_STATE_DATA_CONT state. + + if (outSize == 0 && p->state != LZMA2_STATE_FINISHED) + { + // checkFinishBlock is true. So we expect that block must be finished, + // We can return LZMA_STATUS_NOT_SPECIFIED or LZMA_STATUS_NOT_FINISHED here + // break; + return (ELzma2ParseStatus)LZMA_STATUS_NOT_FINISHED; + } + + if (p->state == LZMA2_STATE_DATA) + return LZMA2_PARSE_STATUS_NEW_CHUNK; + + continue; + } + + if (outSize == 0) + return (ELzma2ParseStatus)LZMA_STATUS_NOT_FINISHED; + + { + SizeT inCur = inSize - *srcLen; + + if (LZMA2_IS_UNCOMPRESSED_STATE(p)) + { + if (inCur == 0) + return (ELzma2ParseStatus)LZMA_STATUS_NEEDS_MORE_INPUT; + if (inCur > p->unpackSize) + inCur = p->unpackSize; + if (inCur > outSize) + inCur = outSize; + p->decoder.dicPos += inCur; + src += inCur; + *srcLen += inCur; + outSize -= inCur; + p->unpackSize -= (UInt32)inCur; + p->state = (p->unpackSize == 0) ? LZMA2_STATE_CONTROL : LZMA2_STATE_DATA_CONT; + } + else + { + p->isExtraMode = True; + + if (inCur == 0) + { + if (p->packSize != 0) + return (ELzma2ParseStatus)LZMA_STATUS_NEEDS_MORE_INPUT; + } + else if (p->state == LZMA2_STATE_DATA) + { + p->state = LZMA2_STATE_DATA_CONT; + if (*src != 0) + { + // first byte of lzma chunk must be Zero + *srcLen += 1; + p->packSize--; + break; + } + } + + if (inCur > p->packSize) + inCur = (SizeT)p->packSize; + + src += inCur; + *srcLen += inCur; + p->packSize -= (UInt32)inCur; + + if (p->packSize == 0) + { + SizeT rem = outSize; + if (rem > p->unpackSize) + rem = p->unpackSize; + p->decoder.dicPos += rem; + p->unpackSize -= (UInt32)rem; + outSize -= rem; + if (p->unpackSize == 0) + p->state = LZMA2_STATE_CONTROL; + } + } + } + } + + p->state = LZMA2_STATE_ERROR; + return (ELzma2ParseStatus)LZMA_STATUS_NOT_SPECIFIED; +} + + + + +SRes Lzma2Dec_DecodeToBuf(CLzma2Dec *p, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status) +{ + SizeT outSize = *destLen, inSize = *srcLen; + *srcLen = *destLen = 0; + + for (;;) + { + SizeT inCur = inSize, outCur, dicPos; + ELzmaFinishMode curFinishMode; + SRes res; + + if (p->decoder.dicPos == p->decoder.dicBufSize) + p->decoder.dicPos = 0; + dicPos = p->decoder.dicPos; + curFinishMode = LZMA_FINISH_ANY; + outCur = p->decoder.dicBufSize - dicPos; + + if (outCur >= outSize) + { + outCur = outSize; + curFinishMode = finishMode; + } + + res = Lzma2Dec_DecodeToDic(p, dicPos + outCur, src, &inCur, curFinishMode, status); + + src += inCur; + inSize -= inCur; + *srcLen += inCur; + outCur = p->decoder.dicPos - dicPos; + memcpy(dest, p->decoder.dic + dicPos, outCur); + dest += outCur; + outSize -= outCur; + *destLen += outCur; + if (res != 0) + return res; + if (outCur == 0 || outSize == 0) + return SZ_OK; + } +} + + +SRes Lzma2Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + Byte prop, ELzmaFinishMode finishMode, ELzmaStatus *status, ISzAllocPtr alloc) +{ + CLzma2Dec p; + SRes res; + SizeT outSize = *destLen, inSize = *srcLen; + *destLen = *srcLen = 0; + *status = LZMA_STATUS_NOT_SPECIFIED; + Lzma2Dec_Construct(&p); + RINOK(Lzma2Dec_AllocateProbs(&p, prop, alloc)); + p.decoder.dic = dest; + p.decoder.dicBufSize = outSize; + Lzma2Dec_Init(&p); + *srcLen = inSize; + res = Lzma2Dec_DecodeToDic(&p, outSize, src, srcLen, finishMode, status); + *destLen = p.decoder.dicPos; + if (res == SZ_OK && *status == LZMA_STATUS_NEEDS_MORE_INPUT) + res = SZ_ERROR_INPUT_EOF; + Lzma2Dec_FreeProbs(&p, alloc); + return res; +} diff --git a/deps/LZMA-SDK/C/Lzma2Dec.h b/deps/LZMA-SDK/C/Lzma2Dec.h index da5038725..b8ddeac89 100644 --- a/deps/LZMA-SDK/C/Lzma2Dec.h +++ b/deps/LZMA-SDK/C/Lzma2Dec.h @@ -1,120 +1,120 @@ -/* Lzma2Dec.h -- LZMA2 Decoder -2018-02-19 : Igor Pavlov : Public domain */ - -#ifndef __LZMA2_DEC_H -#define __LZMA2_DEC_H - -#include "LzmaDec.h" - -EXTERN_C_BEGIN - -/* ---------- State Interface ---------- */ - -typedef struct -{ - unsigned state; - Byte control; - Byte needInitLevel; - Byte isExtraMode; - Byte _pad_; - UInt32 packSize; - UInt32 unpackSize; - CLzmaDec decoder; -} CLzma2Dec; - -#define Lzma2Dec_Construct(p) LzmaDec_Construct(&(p)->decoder) -#define Lzma2Dec_FreeProbs(p, alloc) LzmaDec_FreeProbs(&(p)->decoder, alloc) -#define Lzma2Dec_Free(p, alloc) LzmaDec_Free(&(p)->decoder, alloc) - -SRes Lzma2Dec_AllocateProbs(CLzma2Dec *p, Byte prop, ISzAllocPtr alloc); -SRes Lzma2Dec_Allocate(CLzma2Dec *p, Byte prop, ISzAllocPtr alloc); -void Lzma2Dec_Init(CLzma2Dec *p); - -/* -finishMode: - It has meaning only if the decoding reaches output limit (*destLen or dicLimit). - LZMA_FINISH_ANY - use smallest number of input bytes - LZMA_FINISH_END - read EndOfStream marker after decoding - -Returns: - SZ_OK - status: - LZMA_STATUS_FINISHED_WITH_MARK - LZMA_STATUS_NOT_FINISHED - LZMA_STATUS_NEEDS_MORE_INPUT - SZ_ERROR_DATA - Data error -*/ - -SRes Lzma2Dec_DecodeToDic(CLzma2Dec *p, SizeT dicLimit, - const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); - -SRes Lzma2Dec_DecodeToBuf(CLzma2Dec *p, Byte *dest, SizeT *destLen, - const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); - - -/* ---------- LZMA2 block and chunk parsing ---------- */ - -/* -Lzma2Dec_Parse() parses compressed data stream up to next independent block or next chunk data. -It can return LZMA_STATUS_* code or LZMA2_PARSE_STATUS_* code: - - LZMA2_PARSE_STATUS_NEW_BLOCK - there is new block, and 1 additional byte (control byte of next block header) was read from input. - - LZMA2_PARSE_STATUS_NEW_CHUNK - there is new chunk, and only lzma2 header of new chunk was read. - CLzma2Dec::unpackSize contains unpack size of that chunk -*/ - -typedef enum -{ -/* - LZMA_STATUS_NOT_SPECIFIED // data error - LZMA_STATUS_FINISHED_WITH_MARK - LZMA_STATUS_NOT_FINISHED // - LZMA_STATUS_NEEDS_MORE_INPUT - LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK // unused -*/ - LZMA2_PARSE_STATUS_NEW_BLOCK = LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK + 1, - LZMA2_PARSE_STATUS_NEW_CHUNK -} ELzma2ParseStatus; - -ELzma2ParseStatus Lzma2Dec_Parse(CLzma2Dec *p, - SizeT outSize, // output size - const Byte *src, SizeT *srcLen, - int checkFinishBlock // set (checkFinishBlock = 1), if it must read full input data, if decoder.dicPos reaches blockMax position. - ); - -/* -LZMA2 parser doesn't decode LZMA chunks, so we must read - full input LZMA chunk to decode some part of LZMA chunk. - -Lzma2Dec_GetUnpackExtra() returns the value that shows - max possible number of output bytes that can be output by decoder - at current input positon. -*/ - -#define Lzma2Dec_GetUnpackExtra(p) ((p)->isExtraMode ? (p)->unpackSize : 0); - - -/* ---------- One Call Interface ---------- */ - -/* -finishMode: - It has meaning only if the decoding reaches output limit (*destLen). - LZMA_FINISH_ANY - use smallest number of input bytes - LZMA_FINISH_END - read EndOfStream marker after decoding - -Returns: - SZ_OK - status: - LZMA_STATUS_FINISHED_WITH_MARK - LZMA_STATUS_NOT_FINISHED - SZ_ERROR_DATA - Data error - SZ_ERROR_MEM - Memory allocation error - SZ_ERROR_UNSUPPORTED - Unsupported properties - SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src). -*/ - -SRes Lzma2Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, - Byte prop, ELzmaFinishMode finishMode, ELzmaStatus *status, ISzAllocPtr alloc); - -EXTERN_C_END - -#endif +/* Lzma2Dec.h -- LZMA2 Decoder +2018-02-19 : Igor Pavlov : Public domain */ + +#ifndef __LZMA2_DEC_H +#define __LZMA2_DEC_H + +#include "LzmaDec.h" + +EXTERN_C_BEGIN + +/* ---------- State Interface ---------- */ + +typedef struct +{ + unsigned state; + Byte control; + Byte needInitLevel; + Byte isExtraMode; + Byte _pad_; + UInt32 packSize; + UInt32 unpackSize; + CLzmaDec decoder; +} CLzma2Dec; + +#define Lzma2Dec_Construct(p) LzmaDec_Construct(&(p)->decoder) +#define Lzma2Dec_FreeProbs(p, alloc) LzmaDec_FreeProbs(&(p)->decoder, alloc) +#define Lzma2Dec_Free(p, alloc) LzmaDec_Free(&(p)->decoder, alloc) + +SRes Lzma2Dec_AllocateProbs(CLzma2Dec *p, Byte prop, ISzAllocPtr alloc); +SRes Lzma2Dec_Allocate(CLzma2Dec *p, Byte prop, ISzAllocPtr alloc); +void Lzma2Dec_Init(CLzma2Dec *p); + +/* +finishMode: + It has meaning only if the decoding reaches output limit (*destLen or dicLimit). + LZMA_FINISH_ANY - use smallest number of input bytes + LZMA_FINISH_END - read EndOfStream marker after decoding + +Returns: + SZ_OK + status: + LZMA_STATUS_FINISHED_WITH_MARK + LZMA_STATUS_NOT_FINISHED + LZMA_STATUS_NEEDS_MORE_INPUT + SZ_ERROR_DATA - Data error +*/ + +SRes Lzma2Dec_DecodeToDic(CLzma2Dec *p, SizeT dicLimit, + const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); + +SRes Lzma2Dec_DecodeToBuf(CLzma2Dec *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); + + +/* ---------- LZMA2 block and chunk parsing ---------- */ + +/* +Lzma2Dec_Parse() parses compressed data stream up to next independent block or next chunk data. +It can return LZMA_STATUS_* code or LZMA2_PARSE_STATUS_* code: + - LZMA2_PARSE_STATUS_NEW_BLOCK - there is new block, and 1 additional byte (control byte of next block header) was read from input. + - LZMA2_PARSE_STATUS_NEW_CHUNK - there is new chunk, and only lzma2 header of new chunk was read. + CLzma2Dec::unpackSize contains unpack size of that chunk +*/ + +typedef enum +{ +/* + LZMA_STATUS_NOT_SPECIFIED // data error + LZMA_STATUS_FINISHED_WITH_MARK + LZMA_STATUS_NOT_FINISHED // + LZMA_STATUS_NEEDS_MORE_INPUT + LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK // unused +*/ + LZMA2_PARSE_STATUS_NEW_BLOCK = LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK + 1, + LZMA2_PARSE_STATUS_NEW_CHUNK +} ELzma2ParseStatus; + +ELzma2ParseStatus Lzma2Dec_Parse(CLzma2Dec *p, + SizeT outSize, // output size + const Byte *src, SizeT *srcLen, + int checkFinishBlock // set (checkFinishBlock = 1), if it must read full input data, if decoder.dicPos reaches blockMax position. + ); + +/* +LZMA2 parser doesn't decode LZMA chunks, so we must read + full input LZMA chunk to decode some part of LZMA chunk. + +Lzma2Dec_GetUnpackExtra() returns the value that shows + max possible number of output bytes that can be output by decoder + at current input positon. +*/ + +#define Lzma2Dec_GetUnpackExtra(p) ((p)->isExtraMode ? (p)->unpackSize : 0); + + +/* ---------- One Call Interface ---------- */ + +/* +finishMode: + It has meaning only if the decoding reaches output limit (*destLen). + LZMA_FINISH_ANY - use smallest number of input bytes + LZMA_FINISH_END - read EndOfStream marker after decoding + +Returns: + SZ_OK + status: + LZMA_STATUS_FINISHED_WITH_MARK + LZMA_STATUS_NOT_FINISHED + SZ_ERROR_DATA - Data error + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_UNSUPPORTED - Unsupported properties + SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src). +*/ + +SRes Lzma2Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + Byte prop, ELzmaFinishMode finishMode, ELzmaStatus *status, ISzAllocPtr alloc); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/Lzma2DecMt.c b/deps/LZMA-SDK/C/Lzma2DecMt.c index 252b5be49..9f1dc52ba 100644 --- a/deps/LZMA-SDK/C/Lzma2DecMt.c +++ b/deps/LZMA-SDK/C/Lzma2DecMt.c @@ -1,1090 +1,1090 @@ -/* Lzma2DecMt.c -- LZMA2 Decoder Multi-thread -2021-04-01 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -// #define SHOW_DEBUG_INFO - -// #define _7ZIP_ST - -#ifdef SHOW_DEBUG_INFO -#include -#endif - -#ifndef _7ZIP_ST -#ifdef SHOW_DEBUG_INFO -#define PRF(x) x -#else -#define PRF(x) -#endif -#define PRF_STR(s) PRF(printf("\n" s "\n")) -#define PRF_STR_INT_2(s, d1, d2) PRF(printf("\n" s " %d %d\n", (unsigned)d1, (unsigned)d2)) -#endif - -#include "Alloc.h" - -#include "Lzma2Dec.h" -#include "Lzma2DecMt.h" - -#ifndef _7ZIP_ST -#include "MtDec.h" - -#define LZMA2DECMT_OUT_BLOCK_MAX_DEFAULT (1 << 28) -#endif - - -void Lzma2DecMtProps_Init(CLzma2DecMtProps *p) -{ - p->inBufSize_ST = 1 << 20; - p->outStep_ST = 1 << 20; - - #ifndef _7ZIP_ST - p->numThreads = 1; - p->inBufSize_MT = 1 << 18; - p->outBlockMax = LZMA2DECMT_OUT_BLOCK_MAX_DEFAULT; - p->inBlockMax = p->outBlockMax + p->outBlockMax / 16; - #endif -} - - - -#ifndef _7ZIP_ST - -/* ---------- CLzma2DecMtThread ---------- */ - -typedef struct -{ - CLzma2Dec dec; - Byte dec_created; - Byte needInit; - - Byte *outBuf; - size_t outBufSize; - - EMtDecParseState state; - ELzma2ParseStatus parseStatus; - - size_t inPreSize; - size_t outPreSize; - - size_t inCodeSize; - size_t outCodeSize; - SRes codeRes; - - CAlignOffsetAlloc alloc; - - Byte mtPad[1 << 7]; -} CLzma2DecMtThread; - -#endif - - -/* ---------- CLzma2DecMt ---------- */ - -typedef struct -{ - // ISzAllocPtr alloc; - ISzAllocPtr allocMid; - - CAlignOffsetAlloc alignOffsetAlloc; - CLzma2DecMtProps props; - Byte prop; - - ISeqInStream *inStream; - ISeqOutStream *outStream; - ICompressProgress *progress; - - BoolInt finishMode; - BoolInt outSize_Defined; - UInt64 outSize; - - UInt64 outProcessed; - UInt64 inProcessed; - BoolInt readWasFinished; - SRes readRes; - - Byte *inBuf; - size_t inBufSize; - Byte dec_created; - CLzma2Dec dec; - - size_t inPos; - size_t inLim; - - #ifndef _7ZIP_ST - UInt64 outProcessed_Parse; - BoolInt mtc_WasConstructed; - CMtDec mtc; - CLzma2DecMtThread coders[MTDEC__THREADS_MAX]; - #endif - -} CLzma2DecMt; - - - -CLzma2DecMtHandle Lzma2DecMt_Create(ISzAllocPtr alloc, ISzAllocPtr allocMid) -{ - CLzma2DecMt *p = (CLzma2DecMt *)ISzAlloc_Alloc(alloc, sizeof(CLzma2DecMt)); - if (!p) - return NULL; - - // p->alloc = alloc; - p->allocMid = allocMid; - - AlignOffsetAlloc_CreateVTable(&p->alignOffsetAlloc); - p->alignOffsetAlloc.numAlignBits = 7; - p->alignOffsetAlloc.offset = 0; - p->alignOffsetAlloc.baseAlloc = alloc; - - p->inBuf = NULL; - p->inBufSize = 0; - p->dec_created = False; - - // Lzma2DecMtProps_Init(&p->props); - - #ifndef _7ZIP_ST - p->mtc_WasConstructed = False; - { - unsigned i; - for (i = 0; i < MTDEC__THREADS_MAX; i++) - { - CLzma2DecMtThread *t = &p->coders[i]; - t->dec_created = False; - t->outBuf = NULL; - t->outBufSize = 0; - } - } - #endif - - return p; -} - - -#ifndef _7ZIP_ST - -static void Lzma2DecMt_FreeOutBufs(CLzma2DecMt *p) -{ - unsigned i; - for (i = 0; i < MTDEC__THREADS_MAX; i++) - { - CLzma2DecMtThread *t = &p->coders[i]; - if (t->outBuf) - { - ISzAlloc_Free(p->allocMid, t->outBuf); - t->outBuf = NULL; - t->outBufSize = 0; - } - } -} - -#endif - - -static void Lzma2DecMt_FreeSt(CLzma2DecMt *p) -{ - if (p->dec_created) - { - Lzma2Dec_Free(&p->dec, &p->alignOffsetAlloc.vt); - p->dec_created = False; - } - if (p->inBuf) - { - ISzAlloc_Free(p->allocMid, p->inBuf); - p->inBuf = NULL; - } - p->inBufSize = 0; -} - - -void Lzma2DecMt_Destroy(CLzma2DecMtHandle pp) -{ - CLzma2DecMt *p = (CLzma2DecMt *)pp; - - Lzma2DecMt_FreeSt(p); - - #ifndef _7ZIP_ST - - if (p->mtc_WasConstructed) - { - MtDec_Destruct(&p->mtc); - p->mtc_WasConstructed = False; - } - { - unsigned i; - for (i = 0; i < MTDEC__THREADS_MAX; i++) - { - CLzma2DecMtThread *t = &p->coders[i]; - if (t->dec_created) - { - // we don't need to free dict here - Lzma2Dec_FreeProbs(&t->dec, &t->alloc.vt); // p->alloc !!! - t->dec_created = False; - } - } - } - Lzma2DecMt_FreeOutBufs(p); - - #endif - - ISzAlloc_Free(p->alignOffsetAlloc.baseAlloc, pp); -} - - - -#ifndef _7ZIP_ST - -static void Lzma2DecMt_MtCallback_Parse(void *obj, unsigned coderIndex, CMtDecCallbackInfo *cc) -{ - CLzma2DecMt *me = (CLzma2DecMt *)obj; - CLzma2DecMtThread *t = &me->coders[coderIndex]; - - PRF_STR_INT_2("Parse", coderIndex, cc->srcSize); - - cc->state = MTDEC_PARSE_CONTINUE; - - if (cc->startCall) - { - if (!t->dec_created) - { - Lzma2Dec_Construct(&t->dec); - t->dec_created = True; - AlignOffsetAlloc_CreateVTable(&t->alloc); - { - /* (1 << 12) is expected size of one way in data cache. - We optimize alignment for cache line size of 128 bytes and smaller */ - const unsigned kNumAlignBits = 12; - const unsigned kNumCacheLineBits = 7; /* <= kNumAlignBits */ - t->alloc.numAlignBits = kNumAlignBits; - t->alloc.offset = ((UInt32)coderIndex * (((unsigned)1 << 11) + (1 << 8) + (1 << 6))) & (((unsigned)1 << kNumAlignBits) - ((unsigned)1 << kNumCacheLineBits)); - t->alloc.baseAlloc = me->alignOffsetAlloc.baseAlloc; - } - } - Lzma2Dec_Init(&t->dec); - - t->inPreSize = 0; - t->outPreSize = 0; - // t->blockWasFinished = False; - // t->finishedWithMark = False; - t->parseStatus = (ELzma2ParseStatus)LZMA_STATUS_NOT_SPECIFIED; - t->state = MTDEC_PARSE_CONTINUE; - - t->inCodeSize = 0; - t->outCodeSize = 0; - t->codeRes = SZ_OK; - - // (cc->srcSize == 0) is allowed - } - - { - ELzma2ParseStatus status; - BoolInt overflow; - UInt32 unpackRem = 0; - - int checkFinishBlock = True; - size_t limit = me->props.outBlockMax; - if (me->outSize_Defined) - { - UInt64 rem = me->outSize - me->outProcessed_Parse; - if (limit >= rem) - { - limit = (size_t)rem; - if (!me->finishMode) - checkFinishBlock = False; - } - } - - // checkFinishBlock = False, if we want to decode partial data - // that must be finished at position <= outBlockMax. - - { - const SizeT srcOrig = cc->srcSize; - SizeT srcSize_Point = 0; - SizeT dicPos_Point = 0; - - cc->srcSize = 0; - overflow = False; - - for (;;) - { - SizeT srcCur = srcOrig - cc->srcSize; - - status = Lzma2Dec_Parse(&t->dec, - limit - t->dec.decoder.dicPos, - cc->src + cc->srcSize, &srcCur, - checkFinishBlock); - - cc->srcSize += srcCur; - - if (status == LZMA2_PARSE_STATUS_NEW_CHUNK) - { - if (t->dec.unpackSize > me->props.outBlockMax - t->dec.decoder.dicPos) - { - overflow = True; - break; - } - continue; - } - - if (status == LZMA2_PARSE_STATUS_NEW_BLOCK) - { - if (t->dec.decoder.dicPos == 0) - continue; - // we decode small blocks in one thread - if (t->dec.decoder.dicPos >= (1 << 14)) - break; - dicPos_Point = t->dec.decoder.dicPos; - srcSize_Point = cc->srcSize; - continue; - } - - if ((int)status == LZMA_STATUS_NOT_FINISHED && checkFinishBlock - // && limit == t->dec.decoder.dicPos - // && limit == me->props.outBlockMax - ) - { - overflow = True; - break; - } - - unpackRem = Lzma2Dec_GetUnpackExtra(&t->dec); - break; - } - - if (dicPos_Point != 0 - && (int)status != LZMA2_PARSE_STATUS_NEW_BLOCK - && (int)status != LZMA_STATUS_FINISHED_WITH_MARK - && (int)status != LZMA_STATUS_NOT_SPECIFIED) - { - // we revert to latest newBlock state - status = LZMA2_PARSE_STATUS_NEW_BLOCK; - unpackRem = 0; - t->dec.decoder.dicPos = dicPos_Point; - cc->srcSize = srcSize_Point; - overflow = False; - } - } - - t->inPreSize += cc->srcSize; - t->parseStatus = status; - - if (overflow) - cc->state = MTDEC_PARSE_OVERFLOW; - else - { - size_t dicPos = t->dec.decoder.dicPos; - - if ((int)status != LZMA_STATUS_NEEDS_MORE_INPUT) - { - if (status == LZMA2_PARSE_STATUS_NEW_BLOCK) - { - cc->state = MTDEC_PARSE_NEW; - cc->srcSize--; // we don't need control byte of next block - t->inPreSize--; - } - else - { - cc->state = MTDEC_PARSE_END; - if ((int)status != LZMA_STATUS_FINISHED_WITH_MARK) - { - // (status == LZMA_STATUS_NOT_SPECIFIED) - // (status == LZMA_STATUS_NOT_FINISHED) - if (unpackRem != 0) - { - /* we also reserve space for max possible number of output bytes of current LZMA chunk */ - SizeT rem = limit - dicPos; - if (rem > unpackRem) - rem = unpackRem; - dicPos += rem; - } - } - } - - me->outProcessed_Parse += dicPos; - } - - cc->outPos = dicPos; - t->outPreSize = (size_t)dicPos; - } - - t->state = cc->state; - return; - } -} - - -static SRes Lzma2DecMt_MtCallback_PreCode(void *pp, unsigned coderIndex) -{ - CLzma2DecMt *me = (CLzma2DecMt *)pp; - CLzma2DecMtThread *t = &me->coders[coderIndex]; - Byte *dest = t->outBuf; - - if (t->inPreSize == 0) - { - t->codeRes = SZ_ERROR_DATA; - return t->codeRes; - } - - if (!dest || t->outBufSize < t->outPreSize) - { - if (dest) - { - ISzAlloc_Free(me->allocMid, dest); - t->outBuf = NULL; - t->outBufSize = 0; - } - - dest = (Byte *)ISzAlloc_Alloc(me->allocMid, t->outPreSize - // + (1 << 28) - ); - // Sleep(200); - if (!dest) - return SZ_ERROR_MEM; - t->outBuf = dest; - t->outBufSize = t->outPreSize; - } - - t->dec.decoder.dic = dest; - t->dec.decoder.dicBufSize = t->outPreSize; - - t->needInit = True; - - return Lzma2Dec_AllocateProbs(&t->dec, me->prop, &t->alloc.vt); // alloc.vt -} - - -static SRes Lzma2DecMt_MtCallback_Code(void *pp, unsigned coderIndex, - const Byte *src, size_t srcSize, int srcFinished, - // int finished, int blockFinished, - UInt64 *inCodePos, UInt64 *outCodePos, int *stop) -{ - CLzma2DecMt *me = (CLzma2DecMt *)pp; - CLzma2DecMtThread *t = &me->coders[coderIndex]; - - UNUSED_VAR(srcFinished) - - PRF_STR_INT_2("Code", coderIndex, srcSize); - - *inCodePos = t->inCodeSize; - *outCodePos = 0; - *stop = True; - - if (t->needInit) - { - Lzma2Dec_Init(&t->dec); - t->needInit = False; - } - - { - ELzmaStatus status; - size_t srcProcessed = srcSize; - BoolInt blockWasFinished = - ((int)t->parseStatus == LZMA_STATUS_FINISHED_WITH_MARK - || t->parseStatus == LZMA2_PARSE_STATUS_NEW_BLOCK); - - SRes res = Lzma2Dec_DecodeToDic(&t->dec, - t->outPreSize, - src, &srcProcessed, - blockWasFinished ? LZMA_FINISH_END : LZMA_FINISH_ANY, - &status); - - t->codeRes = res; - - t->inCodeSize += srcProcessed; - *inCodePos = t->inCodeSize; - t->outCodeSize = t->dec.decoder.dicPos; - *outCodePos = t->dec.decoder.dicPos; - - if (res != SZ_OK) - return res; - - if (srcProcessed == srcSize) - *stop = False; - - if (blockWasFinished) - { - if (srcSize != srcProcessed) - return SZ_ERROR_FAIL; - - if (t->inPreSize == t->inCodeSize) - { - if (t->outPreSize != t->outCodeSize) - return SZ_ERROR_FAIL; - *stop = True; - } - } - else - { - if (t->outPreSize == t->outCodeSize) - *stop = True; - } - - return SZ_OK; - } -} - - -#define LZMA2DECMT_STREAM_WRITE_STEP (1 << 24) - -static SRes Lzma2DecMt_MtCallback_Write(void *pp, unsigned coderIndex, - BoolInt needWriteToStream, - const Byte *src, size_t srcSize, BoolInt isCross, - BoolInt *needContinue, BoolInt *canRecode) -{ - CLzma2DecMt *me = (CLzma2DecMt *)pp; - const CLzma2DecMtThread *t = &me->coders[coderIndex]; - size_t size = t->outCodeSize; - const Byte *data = t->outBuf; - BoolInt needContinue2 = True; - - UNUSED_VAR(src) - UNUSED_VAR(srcSize) - UNUSED_VAR(isCross) - - PRF_STR_INT_2("Write", coderIndex, srcSize); - - *needContinue = False; - *canRecode = True; - - if ( - // t->parseStatus == LZMA_STATUS_FINISHED_WITH_MARK - t->state == MTDEC_PARSE_OVERFLOW - || t->state == MTDEC_PARSE_END) - needContinue2 = False; - - - if (!needWriteToStream) - return SZ_OK; - - me->mtc.inProcessed += t->inCodeSize; - - if (t->codeRes == SZ_OK) - if ((int)t->parseStatus == LZMA_STATUS_FINISHED_WITH_MARK - || t->parseStatus == LZMA2_PARSE_STATUS_NEW_BLOCK) - if (t->outPreSize != t->outCodeSize - || t->inPreSize != t->inCodeSize) - return SZ_ERROR_FAIL; - - *canRecode = False; - - if (me->outStream) - { - for (;;) - { - size_t cur = size; - size_t written; - if (cur > LZMA2DECMT_STREAM_WRITE_STEP) - cur = LZMA2DECMT_STREAM_WRITE_STEP; - - written = ISeqOutStream_Write(me->outStream, data, cur); - - me->outProcessed += written; - // me->mtc.writtenTotal += written; - if (written != cur) - return SZ_ERROR_WRITE; - data += cur; - size -= cur; - if (size == 0) - { - *needContinue = needContinue2; - return SZ_OK; - } - RINOK(MtProgress_ProgressAdd(&me->mtc.mtProgress, 0, 0)); - } - } - - return SZ_ERROR_FAIL; - /* - if (size > me->outBufSize) - return SZ_ERROR_OUTPUT_EOF; - memcpy(me->outBuf, data, size); - me->outBufSize -= size; - me->outBuf += size; - *needContinue = needContinue2; - return SZ_OK; - */ -} - -#endif - - -static SRes Lzma2Dec_Prepare_ST(CLzma2DecMt *p) -{ - if (!p->dec_created) - { - Lzma2Dec_Construct(&p->dec); - p->dec_created = True; - } - - RINOK(Lzma2Dec_Allocate(&p->dec, p->prop, &p->alignOffsetAlloc.vt)); - - if (!p->inBuf || p->inBufSize != p->props.inBufSize_ST) - { - ISzAlloc_Free(p->allocMid, p->inBuf); - p->inBufSize = 0; - p->inBuf = (Byte *)ISzAlloc_Alloc(p->allocMid, p->props.inBufSize_ST); - if (!p->inBuf) - return SZ_ERROR_MEM; - p->inBufSize = p->props.inBufSize_ST; - } - - Lzma2Dec_Init(&p->dec); - - return SZ_OK; -} - - -static SRes Lzma2Dec_Decode_ST(CLzma2DecMt *p - #ifndef _7ZIP_ST - , BoolInt tMode - #endif - ) -{ - SizeT wrPos; - size_t inPos, inLim; - const Byte *inData; - UInt64 inPrev, outPrev; - - CLzma2Dec *dec; - - #ifndef _7ZIP_ST - if (tMode) - { - Lzma2DecMt_FreeOutBufs(p); - tMode = MtDec_PrepareRead(&p->mtc); - } - #endif - - RINOK(Lzma2Dec_Prepare_ST(p)); - - dec = &p->dec; - - inPrev = p->inProcessed; - outPrev = p->outProcessed; - - inPos = 0; - inLim = 0; - inData = NULL; - wrPos = dec->decoder.dicPos; - - for (;;) - { - SizeT dicPos; - SizeT size; - ELzmaFinishMode finishMode; - SizeT inProcessed; - ELzmaStatus status; - SRes res; - - SizeT outProcessed; - BoolInt outFinished; - BoolInt needStop; - - if (inPos == inLim) - { - #ifndef _7ZIP_ST - if (tMode) - { - inData = MtDec_Read(&p->mtc, &inLim); - inPos = 0; - if (inData) - continue; - tMode = False; - inLim = 0; - } - #endif - - if (!p->readWasFinished) - { - inPos = 0; - inLim = p->inBufSize; - inData = p->inBuf; - p->readRes = ISeqInStream_Read(p->inStream, (void *)(p->inBuf), &inLim); - // p->readProcessed += inLim; - // inLim -= 5; p->readWasFinished = True; // for test - if (inLim == 0 || p->readRes != SZ_OK) - p->readWasFinished = True; - } - } - - dicPos = dec->decoder.dicPos; - { - SizeT next = dec->decoder.dicBufSize; - if (next - wrPos > p->props.outStep_ST) - next = wrPos + p->props.outStep_ST; - size = next - dicPos; - } - - finishMode = LZMA_FINISH_ANY; - if (p->outSize_Defined) - { - const UInt64 rem = p->outSize - p->outProcessed; - if (size >= rem) - { - size = (SizeT)rem; - if (p->finishMode) - finishMode = LZMA_FINISH_END; - } - } - - inProcessed = inLim - inPos; - - res = Lzma2Dec_DecodeToDic(dec, dicPos + size, inData + inPos, &inProcessed, finishMode, &status); - - inPos += inProcessed; - p->inProcessed += inProcessed; - outProcessed = dec->decoder.dicPos - dicPos; - p->outProcessed += outProcessed; - - outFinished = (p->outSize_Defined && p->outSize <= p->outProcessed); - - needStop = (res != SZ_OK - || (inProcessed == 0 && outProcessed == 0) - || status == LZMA_STATUS_FINISHED_WITH_MARK - || (!p->finishMode && outFinished)); - - if (needStop || outProcessed >= size) - { - SRes res2; - { - size_t writeSize = dec->decoder.dicPos - wrPos; - size_t written = ISeqOutStream_Write(p->outStream, dec->decoder.dic + wrPos, writeSize); - res2 = (written == writeSize) ? SZ_OK : SZ_ERROR_WRITE; - } - - if (dec->decoder.dicPos == dec->decoder.dicBufSize) - dec->decoder.dicPos = 0; - wrPos = dec->decoder.dicPos; - - RINOK(res2); - - if (needStop) - { - if (res != SZ_OK) - return res; - - if (status == LZMA_STATUS_FINISHED_WITH_MARK) - { - if (p->finishMode) - { - if (p->outSize_Defined && p->outSize != p->outProcessed) - return SZ_ERROR_DATA; - } - return SZ_OK; - } - - if (!p->finishMode && outFinished) - return SZ_OK; - - if (status == LZMA_STATUS_NEEDS_MORE_INPUT) - return SZ_ERROR_INPUT_EOF; - - return SZ_ERROR_DATA; - } - } - - if (p->progress) - { - UInt64 inDelta = p->inProcessed - inPrev; - UInt64 outDelta = p->outProcessed - outPrev; - if (inDelta >= (1 << 22) || outDelta >= (1 << 22)) - { - RINOK(ICompressProgress_Progress(p->progress, p->inProcessed, p->outProcessed)); - inPrev = p->inProcessed; - outPrev = p->outProcessed; - } - } - } -} - - - -SRes Lzma2DecMt_Decode(CLzma2DecMtHandle pp, - Byte prop, - const CLzma2DecMtProps *props, - ISeqOutStream *outStream, const UInt64 *outDataSize, int finishMode, - // Byte *outBuf, size_t *outBufSize, - ISeqInStream *inStream, - // const Byte *inData, size_t inDataSize, - UInt64 *inProcessed, - // UInt64 *outProcessed, - int *isMT, - ICompressProgress *progress) -{ - CLzma2DecMt *p = (CLzma2DecMt *)pp; - #ifndef _7ZIP_ST - BoolInt tMode; - #endif - - *inProcessed = 0; - - if (prop > 40) - return SZ_ERROR_UNSUPPORTED; - - p->prop = prop; - p->props = *props; - - p->inStream = inStream; - p->outStream = outStream; - p->progress = progress; - - p->outSize = 0; - p->outSize_Defined = False; - if (outDataSize) - { - p->outSize_Defined = True; - p->outSize = *outDataSize; - } - p->finishMode = finishMode; - - p->outProcessed = 0; - p->inProcessed = 0; - - p->readWasFinished = False; - p->readRes = SZ_OK; - - *isMT = False; - - - #ifndef _7ZIP_ST - - tMode = False; - - // p->mtc.parseRes = SZ_OK; - - // p->mtc.numFilledThreads = 0; - // p->mtc.crossStart = 0; - // p->mtc.crossEnd = 0; - // p->mtc.allocError_for_Read_BlockIndex = 0; - // p->mtc.isAllocError = False; - - if (p->props.numThreads > 1) - { - IMtDecCallback2 vt; - - Lzma2DecMt_FreeSt(p); - - p->outProcessed_Parse = 0; - - if (!p->mtc_WasConstructed) - { - p->mtc_WasConstructed = True; - MtDec_Construct(&p->mtc); - } - - p->mtc.progress = progress; - p->mtc.inStream = inStream; - - // p->outBuf = NULL; - // p->outBufSize = 0; - /* - if (!outStream) - { - // p->outBuf = outBuf; - // p->outBufSize = *outBufSize; - // *outBufSize = 0; - return SZ_ERROR_PARAM; - } - */ - - // p->mtc.inBlockMax = p->props.inBlockMax; - p->mtc.alloc = &p->alignOffsetAlloc.vt; - // p->alignOffsetAlloc.baseAlloc; - // p->mtc.inData = inData; - // p->mtc.inDataSize = inDataSize; - p->mtc.mtCallback = &vt; - p->mtc.mtCallbackObject = p; - - p->mtc.inBufSize = p->props.inBufSize_MT; - - p->mtc.numThreadsMax = p->props.numThreads; - - *isMT = True; - - vt.Parse = Lzma2DecMt_MtCallback_Parse; - vt.PreCode = Lzma2DecMt_MtCallback_PreCode; - vt.Code = Lzma2DecMt_MtCallback_Code; - vt.Write = Lzma2DecMt_MtCallback_Write; - - { - BoolInt needContinue = False; - - SRes res = MtDec_Code(&p->mtc); - - /* - if (!outStream) - *outBufSize = p->outBuf - outBuf; - */ - - *inProcessed = p->mtc.inProcessed; - - needContinue = False; - - if (res == SZ_OK) - { - if (p->mtc.mtProgress.res != SZ_OK) - res = p->mtc.mtProgress.res; - else - needContinue = p->mtc.needContinue; - } - - if (!needContinue) - { - if (res == SZ_OK) - return p->mtc.readRes; - return res; - } - - tMode = True; - p->readRes = p->mtc.readRes; - p->readWasFinished = p->mtc.readWasFinished; - p->inProcessed = p->mtc.inProcessed; - - PRF_STR("----- decoding ST -----"); - } - } - - #endif - - - *isMT = False; - - { - SRes res = Lzma2Dec_Decode_ST(p - #ifndef _7ZIP_ST - , tMode - #endif - ); - - *inProcessed = p->inProcessed; - - // res = SZ_OK; // for test - if (res == SZ_ERROR_INPUT_EOF) - { - if (p->readRes != SZ_OK) - res = p->readRes; - } - else if (res == SZ_OK && p->readRes != SZ_OK) - res = p->readRes; - - /* - #ifndef _7ZIP_ST - if (res == SZ_OK && tMode && p->mtc.parseRes != SZ_OK) - res = p->mtc.parseRes; - #endif - */ - - return res; - } -} - - -/* ---------- Read from CLzma2DecMtHandle Interface ---------- */ - -SRes Lzma2DecMt_Init(CLzma2DecMtHandle pp, - Byte prop, - const CLzma2DecMtProps *props, - const UInt64 *outDataSize, int finishMode, - ISeqInStream *inStream) -{ - CLzma2DecMt *p = (CLzma2DecMt *)pp; - - if (prop > 40) - return SZ_ERROR_UNSUPPORTED; - - p->prop = prop; - p->props = *props; - - p->inStream = inStream; - - p->outSize = 0; - p->outSize_Defined = False; - if (outDataSize) - { - p->outSize_Defined = True; - p->outSize = *outDataSize; - } - p->finishMode = finishMode; - - p->outProcessed = 0; - p->inProcessed = 0; - - p->inPos = 0; - p->inLim = 0; - - return Lzma2Dec_Prepare_ST(p); -} - - -SRes Lzma2DecMt_Read(CLzma2DecMtHandle pp, - Byte *data, size_t *outSize, - UInt64 *inStreamProcessed) -{ - CLzma2DecMt *p = (CLzma2DecMt *)pp; - ELzmaFinishMode finishMode; - SRes readRes; - size_t size = *outSize; - - *outSize = 0; - *inStreamProcessed = 0; - - finishMode = LZMA_FINISH_ANY; - if (p->outSize_Defined) - { - const UInt64 rem = p->outSize - p->outProcessed; - if (size >= rem) - { - size = (size_t)rem; - if (p->finishMode) - finishMode = LZMA_FINISH_END; - } - } - - readRes = SZ_OK; - - for (;;) - { - SizeT inCur; - SizeT outCur; - ELzmaStatus status; - SRes res; - - if (p->inPos == p->inLim && readRes == SZ_OK) - { - p->inPos = 0; - p->inLim = p->props.inBufSize_ST; - readRes = ISeqInStream_Read(p->inStream, p->inBuf, &p->inLim); - } - - inCur = p->inLim - p->inPos; - outCur = size; - - res = Lzma2Dec_DecodeToBuf(&p->dec, data, &outCur, - p->inBuf + p->inPos, &inCur, finishMode, &status); - - p->inPos += inCur; - p->inProcessed += inCur; - *inStreamProcessed += inCur; - p->outProcessed += outCur; - *outSize += outCur; - size -= outCur; - data += outCur; - - if (res != 0) - return res; - - /* - if (status == LZMA_STATUS_FINISHED_WITH_MARK) - return readRes; - - if (size == 0 && status != LZMA_STATUS_NEEDS_MORE_INPUT) - { - if (p->finishMode && p->outSize_Defined && p->outProcessed >= p->outSize) - return SZ_ERROR_DATA; - return readRes; - } - */ - - if (inCur == 0 && outCur == 0) - return readRes; - } -} +/* Lzma2DecMt.c -- LZMA2 Decoder Multi-thread +2021-04-01 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +// #define SHOW_DEBUG_INFO + +// #define _7ZIP_ST + +#ifdef SHOW_DEBUG_INFO +#include +#endif + +#ifndef _7ZIP_ST +#ifdef SHOW_DEBUG_INFO +#define PRF(x) x +#else +#define PRF(x) +#endif +#define PRF_STR(s) PRF(printf("\n" s "\n")) +#define PRF_STR_INT_2(s, d1, d2) PRF(printf("\n" s " %d %d\n", (unsigned)d1, (unsigned)d2)) +#endif + +#include "Alloc.h" + +#include "Lzma2Dec.h" +#include "Lzma2DecMt.h" + +#ifndef _7ZIP_ST +#include "MtDec.h" + +#define LZMA2DECMT_OUT_BLOCK_MAX_DEFAULT (1 << 28) +#endif + + +void Lzma2DecMtProps_Init(CLzma2DecMtProps *p) +{ + p->inBufSize_ST = 1 << 20; + p->outStep_ST = 1 << 20; + + #ifndef _7ZIP_ST + p->numThreads = 1; + p->inBufSize_MT = 1 << 18; + p->outBlockMax = LZMA2DECMT_OUT_BLOCK_MAX_DEFAULT; + p->inBlockMax = p->outBlockMax + p->outBlockMax / 16; + #endif +} + + + +#ifndef _7ZIP_ST + +/* ---------- CLzma2DecMtThread ---------- */ + +typedef struct +{ + CLzma2Dec dec; + Byte dec_created; + Byte needInit; + + Byte *outBuf; + size_t outBufSize; + + EMtDecParseState state; + ELzma2ParseStatus parseStatus; + + size_t inPreSize; + size_t outPreSize; + + size_t inCodeSize; + size_t outCodeSize; + SRes codeRes; + + CAlignOffsetAlloc alloc; + + Byte mtPad[1 << 7]; +} CLzma2DecMtThread; + +#endif + + +/* ---------- CLzma2DecMt ---------- */ + +typedef struct +{ + // ISzAllocPtr alloc; + ISzAllocPtr allocMid; + + CAlignOffsetAlloc alignOffsetAlloc; + CLzma2DecMtProps props; + Byte prop; + + ISeqInStream *inStream; + ISeqOutStream *outStream; + ICompressProgress *progress; + + BoolInt finishMode; + BoolInt outSize_Defined; + UInt64 outSize; + + UInt64 outProcessed; + UInt64 inProcessed; + BoolInt readWasFinished; + SRes readRes; + + Byte *inBuf; + size_t inBufSize; + Byte dec_created; + CLzma2Dec dec; + + size_t inPos; + size_t inLim; + + #ifndef _7ZIP_ST + UInt64 outProcessed_Parse; + BoolInt mtc_WasConstructed; + CMtDec mtc; + CLzma2DecMtThread coders[MTDEC__THREADS_MAX]; + #endif + +} CLzma2DecMt; + + + +CLzma2DecMtHandle Lzma2DecMt_Create(ISzAllocPtr alloc, ISzAllocPtr allocMid) +{ + CLzma2DecMt *p = (CLzma2DecMt *)ISzAlloc_Alloc(alloc, sizeof(CLzma2DecMt)); + if (!p) + return NULL; + + // p->alloc = alloc; + p->allocMid = allocMid; + + AlignOffsetAlloc_CreateVTable(&p->alignOffsetAlloc); + p->alignOffsetAlloc.numAlignBits = 7; + p->alignOffsetAlloc.offset = 0; + p->alignOffsetAlloc.baseAlloc = alloc; + + p->inBuf = NULL; + p->inBufSize = 0; + p->dec_created = False; + + // Lzma2DecMtProps_Init(&p->props); + + #ifndef _7ZIP_ST + p->mtc_WasConstructed = False; + { + unsigned i; + for (i = 0; i < MTDEC__THREADS_MAX; i++) + { + CLzma2DecMtThread *t = &p->coders[i]; + t->dec_created = False; + t->outBuf = NULL; + t->outBufSize = 0; + } + } + #endif + + return p; +} + + +#ifndef _7ZIP_ST + +static void Lzma2DecMt_FreeOutBufs(CLzma2DecMt *p) +{ + unsigned i; + for (i = 0; i < MTDEC__THREADS_MAX; i++) + { + CLzma2DecMtThread *t = &p->coders[i]; + if (t->outBuf) + { + ISzAlloc_Free(p->allocMid, t->outBuf); + t->outBuf = NULL; + t->outBufSize = 0; + } + } +} + +#endif + + +static void Lzma2DecMt_FreeSt(CLzma2DecMt *p) +{ + if (p->dec_created) + { + Lzma2Dec_Free(&p->dec, &p->alignOffsetAlloc.vt); + p->dec_created = False; + } + if (p->inBuf) + { + ISzAlloc_Free(p->allocMid, p->inBuf); + p->inBuf = NULL; + } + p->inBufSize = 0; +} + + +void Lzma2DecMt_Destroy(CLzma2DecMtHandle pp) +{ + CLzma2DecMt *p = (CLzma2DecMt *)pp; + + Lzma2DecMt_FreeSt(p); + + #ifndef _7ZIP_ST + + if (p->mtc_WasConstructed) + { + MtDec_Destruct(&p->mtc); + p->mtc_WasConstructed = False; + } + { + unsigned i; + for (i = 0; i < MTDEC__THREADS_MAX; i++) + { + CLzma2DecMtThread *t = &p->coders[i]; + if (t->dec_created) + { + // we don't need to free dict here + Lzma2Dec_FreeProbs(&t->dec, &t->alloc.vt); // p->alloc !!! + t->dec_created = False; + } + } + } + Lzma2DecMt_FreeOutBufs(p); + + #endif + + ISzAlloc_Free(p->alignOffsetAlloc.baseAlloc, pp); +} + + + +#ifndef _7ZIP_ST + +static void Lzma2DecMt_MtCallback_Parse(void *obj, unsigned coderIndex, CMtDecCallbackInfo *cc) +{ + CLzma2DecMt *me = (CLzma2DecMt *)obj; + CLzma2DecMtThread *t = &me->coders[coderIndex]; + + PRF_STR_INT_2("Parse", coderIndex, cc->srcSize); + + cc->state = MTDEC_PARSE_CONTINUE; + + if (cc->startCall) + { + if (!t->dec_created) + { + Lzma2Dec_Construct(&t->dec); + t->dec_created = True; + AlignOffsetAlloc_CreateVTable(&t->alloc); + { + /* (1 << 12) is expected size of one way in data cache. + We optimize alignment for cache line size of 128 bytes and smaller */ + const unsigned kNumAlignBits = 12; + const unsigned kNumCacheLineBits = 7; /* <= kNumAlignBits */ + t->alloc.numAlignBits = kNumAlignBits; + t->alloc.offset = ((UInt32)coderIndex * (((unsigned)1 << 11) + (1 << 8) + (1 << 6))) & (((unsigned)1 << kNumAlignBits) - ((unsigned)1 << kNumCacheLineBits)); + t->alloc.baseAlloc = me->alignOffsetAlloc.baseAlloc; + } + } + Lzma2Dec_Init(&t->dec); + + t->inPreSize = 0; + t->outPreSize = 0; + // t->blockWasFinished = False; + // t->finishedWithMark = False; + t->parseStatus = (ELzma2ParseStatus)LZMA_STATUS_NOT_SPECIFIED; + t->state = MTDEC_PARSE_CONTINUE; + + t->inCodeSize = 0; + t->outCodeSize = 0; + t->codeRes = SZ_OK; + + // (cc->srcSize == 0) is allowed + } + + { + ELzma2ParseStatus status; + BoolInt overflow; + UInt32 unpackRem = 0; + + int checkFinishBlock = True; + size_t limit = me->props.outBlockMax; + if (me->outSize_Defined) + { + UInt64 rem = me->outSize - me->outProcessed_Parse; + if (limit >= rem) + { + limit = (size_t)rem; + if (!me->finishMode) + checkFinishBlock = False; + } + } + + // checkFinishBlock = False, if we want to decode partial data + // that must be finished at position <= outBlockMax. + + { + const SizeT srcOrig = cc->srcSize; + SizeT srcSize_Point = 0; + SizeT dicPos_Point = 0; + + cc->srcSize = 0; + overflow = False; + + for (;;) + { + SizeT srcCur = srcOrig - cc->srcSize; + + status = Lzma2Dec_Parse(&t->dec, + limit - t->dec.decoder.dicPos, + cc->src + cc->srcSize, &srcCur, + checkFinishBlock); + + cc->srcSize += srcCur; + + if (status == LZMA2_PARSE_STATUS_NEW_CHUNK) + { + if (t->dec.unpackSize > me->props.outBlockMax - t->dec.decoder.dicPos) + { + overflow = True; + break; + } + continue; + } + + if (status == LZMA2_PARSE_STATUS_NEW_BLOCK) + { + if (t->dec.decoder.dicPos == 0) + continue; + // we decode small blocks in one thread + if (t->dec.decoder.dicPos >= (1 << 14)) + break; + dicPos_Point = t->dec.decoder.dicPos; + srcSize_Point = cc->srcSize; + continue; + } + + if ((int)status == LZMA_STATUS_NOT_FINISHED && checkFinishBlock + // && limit == t->dec.decoder.dicPos + // && limit == me->props.outBlockMax + ) + { + overflow = True; + break; + } + + unpackRem = Lzma2Dec_GetUnpackExtra(&t->dec); + break; + } + + if (dicPos_Point != 0 + && (int)status != LZMA2_PARSE_STATUS_NEW_BLOCK + && (int)status != LZMA_STATUS_FINISHED_WITH_MARK + && (int)status != LZMA_STATUS_NOT_SPECIFIED) + { + // we revert to latest newBlock state + status = LZMA2_PARSE_STATUS_NEW_BLOCK; + unpackRem = 0; + t->dec.decoder.dicPos = dicPos_Point; + cc->srcSize = srcSize_Point; + overflow = False; + } + } + + t->inPreSize += cc->srcSize; + t->parseStatus = status; + + if (overflow) + cc->state = MTDEC_PARSE_OVERFLOW; + else + { + size_t dicPos = t->dec.decoder.dicPos; + + if ((int)status != LZMA_STATUS_NEEDS_MORE_INPUT) + { + if (status == LZMA2_PARSE_STATUS_NEW_BLOCK) + { + cc->state = MTDEC_PARSE_NEW; + cc->srcSize--; // we don't need control byte of next block + t->inPreSize--; + } + else + { + cc->state = MTDEC_PARSE_END; + if ((int)status != LZMA_STATUS_FINISHED_WITH_MARK) + { + // (status == LZMA_STATUS_NOT_SPECIFIED) + // (status == LZMA_STATUS_NOT_FINISHED) + if (unpackRem != 0) + { + /* we also reserve space for max possible number of output bytes of current LZMA chunk */ + SizeT rem = limit - dicPos; + if (rem > unpackRem) + rem = unpackRem; + dicPos += rem; + } + } + } + + me->outProcessed_Parse += dicPos; + } + + cc->outPos = dicPos; + t->outPreSize = (size_t)dicPos; + } + + t->state = cc->state; + return; + } +} + + +static SRes Lzma2DecMt_MtCallback_PreCode(void *pp, unsigned coderIndex) +{ + CLzma2DecMt *me = (CLzma2DecMt *)pp; + CLzma2DecMtThread *t = &me->coders[coderIndex]; + Byte *dest = t->outBuf; + + if (t->inPreSize == 0) + { + t->codeRes = SZ_ERROR_DATA; + return t->codeRes; + } + + if (!dest || t->outBufSize < t->outPreSize) + { + if (dest) + { + ISzAlloc_Free(me->allocMid, dest); + t->outBuf = NULL; + t->outBufSize = 0; + } + + dest = (Byte *)ISzAlloc_Alloc(me->allocMid, t->outPreSize + // + (1 << 28) + ); + // Sleep(200); + if (!dest) + return SZ_ERROR_MEM; + t->outBuf = dest; + t->outBufSize = t->outPreSize; + } + + t->dec.decoder.dic = dest; + t->dec.decoder.dicBufSize = t->outPreSize; + + t->needInit = True; + + return Lzma2Dec_AllocateProbs(&t->dec, me->prop, &t->alloc.vt); // alloc.vt +} + + +static SRes Lzma2DecMt_MtCallback_Code(void *pp, unsigned coderIndex, + const Byte *src, size_t srcSize, int srcFinished, + // int finished, int blockFinished, + UInt64 *inCodePos, UInt64 *outCodePos, int *stop) +{ + CLzma2DecMt *me = (CLzma2DecMt *)pp; + CLzma2DecMtThread *t = &me->coders[coderIndex]; + + UNUSED_VAR(srcFinished) + + PRF_STR_INT_2("Code", coderIndex, srcSize); + + *inCodePos = t->inCodeSize; + *outCodePos = 0; + *stop = True; + + if (t->needInit) + { + Lzma2Dec_Init(&t->dec); + t->needInit = False; + } + + { + ELzmaStatus status; + size_t srcProcessed = srcSize; + BoolInt blockWasFinished = + ((int)t->parseStatus == LZMA_STATUS_FINISHED_WITH_MARK + || t->parseStatus == LZMA2_PARSE_STATUS_NEW_BLOCK); + + SRes res = Lzma2Dec_DecodeToDic(&t->dec, + t->outPreSize, + src, &srcProcessed, + blockWasFinished ? LZMA_FINISH_END : LZMA_FINISH_ANY, + &status); + + t->codeRes = res; + + t->inCodeSize += srcProcessed; + *inCodePos = t->inCodeSize; + t->outCodeSize = t->dec.decoder.dicPos; + *outCodePos = t->dec.decoder.dicPos; + + if (res != SZ_OK) + return res; + + if (srcProcessed == srcSize) + *stop = False; + + if (blockWasFinished) + { + if (srcSize != srcProcessed) + return SZ_ERROR_FAIL; + + if (t->inPreSize == t->inCodeSize) + { + if (t->outPreSize != t->outCodeSize) + return SZ_ERROR_FAIL; + *stop = True; + } + } + else + { + if (t->outPreSize == t->outCodeSize) + *stop = True; + } + + return SZ_OK; + } +} + + +#define LZMA2DECMT_STREAM_WRITE_STEP (1 << 24) + +static SRes Lzma2DecMt_MtCallback_Write(void *pp, unsigned coderIndex, + BoolInt needWriteToStream, + const Byte *src, size_t srcSize, BoolInt isCross, + BoolInt *needContinue, BoolInt *canRecode) +{ + CLzma2DecMt *me = (CLzma2DecMt *)pp; + const CLzma2DecMtThread *t = &me->coders[coderIndex]; + size_t size = t->outCodeSize; + const Byte *data = t->outBuf; + BoolInt needContinue2 = True; + + UNUSED_VAR(src) + UNUSED_VAR(srcSize) + UNUSED_VAR(isCross) + + PRF_STR_INT_2("Write", coderIndex, srcSize); + + *needContinue = False; + *canRecode = True; + + if ( + // t->parseStatus == LZMA_STATUS_FINISHED_WITH_MARK + t->state == MTDEC_PARSE_OVERFLOW + || t->state == MTDEC_PARSE_END) + needContinue2 = False; + + + if (!needWriteToStream) + return SZ_OK; + + me->mtc.inProcessed += t->inCodeSize; + + if (t->codeRes == SZ_OK) + if ((int)t->parseStatus == LZMA_STATUS_FINISHED_WITH_MARK + || t->parseStatus == LZMA2_PARSE_STATUS_NEW_BLOCK) + if (t->outPreSize != t->outCodeSize + || t->inPreSize != t->inCodeSize) + return SZ_ERROR_FAIL; + + *canRecode = False; + + if (me->outStream) + { + for (;;) + { + size_t cur = size; + size_t written; + if (cur > LZMA2DECMT_STREAM_WRITE_STEP) + cur = LZMA2DECMT_STREAM_WRITE_STEP; + + written = ISeqOutStream_Write(me->outStream, data, cur); + + me->outProcessed += written; + // me->mtc.writtenTotal += written; + if (written != cur) + return SZ_ERROR_WRITE; + data += cur; + size -= cur; + if (size == 0) + { + *needContinue = needContinue2; + return SZ_OK; + } + RINOK(MtProgress_ProgressAdd(&me->mtc.mtProgress, 0, 0)); + } + } + + return SZ_ERROR_FAIL; + /* + if (size > me->outBufSize) + return SZ_ERROR_OUTPUT_EOF; + memcpy(me->outBuf, data, size); + me->outBufSize -= size; + me->outBuf += size; + *needContinue = needContinue2; + return SZ_OK; + */ +} + +#endif + + +static SRes Lzma2Dec_Prepare_ST(CLzma2DecMt *p) +{ + if (!p->dec_created) + { + Lzma2Dec_Construct(&p->dec); + p->dec_created = True; + } + + RINOK(Lzma2Dec_Allocate(&p->dec, p->prop, &p->alignOffsetAlloc.vt)); + + if (!p->inBuf || p->inBufSize != p->props.inBufSize_ST) + { + ISzAlloc_Free(p->allocMid, p->inBuf); + p->inBufSize = 0; + p->inBuf = (Byte *)ISzAlloc_Alloc(p->allocMid, p->props.inBufSize_ST); + if (!p->inBuf) + return SZ_ERROR_MEM; + p->inBufSize = p->props.inBufSize_ST; + } + + Lzma2Dec_Init(&p->dec); + + return SZ_OK; +} + + +static SRes Lzma2Dec_Decode_ST(CLzma2DecMt *p + #ifndef _7ZIP_ST + , BoolInt tMode + #endif + ) +{ + SizeT wrPos; + size_t inPos, inLim; + const Byte *inData; + UInt64 inPrev, outPrev; + + CLzma2Dec *dec; + + #ifndef _7ZIP_ST + if (tMode) + { + Lzma2DecMt_FreeOutBufs(p); + tMode = MtDec_PrepareRead(&p->mtc); + } + #endif + + RINOK(Lzma2Dec_Prepare_ST(p)); + + dec = &p->dec; + + inPrev = p->inProcessed; + outPrev = p->outProcessed; + + inPos = 0; + inLim = 0; + inData = NULL; + wrPos = dec->decoder.dicPos; + + for (;;) + { + SizeT dicPos; + SizeT size; + ELzmaFinishMode finishMode; + SizeT inProcessed; + ELzmaStatus status; + SRes res; + + SizeT outProcessed; + BoolInt outFinished; + BoolInt needStop; + + if (inPos == inLim) + { + #ifndef _7ZIP_ST + if (tMode) + { + inData = MtDec_Read(&p->mtc, &inLim); + inPos = 0; + if (inData) + continue; + tMode = False; + inLim = 0; + } + #endif + + if (!p->readWasFinished) + { + inPos = 0; + inLim = p->inBufSize; + inData = p->inBuf; + p->readRes = ISeqInStream_Read(p->inStream, (void *)(p->inBuf), &inLim); + // p->readProcessed += inLim; + // inLim -= 5; p->readWasFinished = True; // for test + if (inLim == 0 || p->readRes != SZ_OK) + p->readWasFinished = True; + } + } + + dicPos = dec->decoder.dicPos; + { + SizeT next = dec->decoder.dicBufSize; + if (next - wrPos > p->props.outStep_ST) + next = wrPos + p->props.outStep_ST; + size = next - dicPos; + } + + finishMode = LZMA_FINISH_ANY; + if (p->outSize_Defined) + { + const UInt64 rem = p->outSize - p->outProcessed; + if (size >= rem) + { + size = (SizeT)rem; + if (p->finishMode) + finishMode = LZMA_FINISH_END; + } + } + + inProcessed = inLim - inPos; + + res = Lzma2Dec_DecodeToDic(dec, dicPos + size, inData + inPos, &inProcessed, finishMode, &status); + + inPos += inProcessed; + p->inProcessed += inProcessed; + outProcessed = dec->decoder.dicPos - dicPos; + p->outProcessed += outProcessed; + + outFinished = (p->outSize_Defined && p->outSize <= p->outProcessed); + + needStop = (res != SZ_OK + || (inProcessed == 0 && outProcessed == 0) + || status == LZMA_STATUS_FINISHED_WITH_MARK + || (!p->finishMode && outFinished)); + + if (needStop || outProcessed >= size) + { + SRes res2; + { + size_t writeSize = dec->decoder.dicPos - wrPos; + size_t written = ISeqOutStream_Write(p->outStream, dec->decoder.dic + wrPos, writeSize); + res2 = (written == writeSize) ? SZ_OK : SZ_ERROR_WRITE; + } + + if (dec->decoder.dicPos == dec->decoder.dicBufSize) + dec->decoder.dicPos = 0; + wrPos = dec->decoder.dicPos; + + RINOK(res2); + + if (needStop) + { + if (res != SZ_OK) + return res; + + if (status == LZMA_STATUS_FINISHED_WITH_MARK) + { + if (p->finishMode) + { + if (p->outSize_Defined && p->outSize != p->outProcessed) + return SZ_ERROR_DATA; + } + return SZ_OK; + } + + if (!p->finishMode && outFinished) + return SZ_OK; + + if (status == LZMA_STATUS_NEEDS_MORE_INPUT) + return SZ_ERROR_INPUT_EOF; + + return SZ_ERROR_DATA; + } + } + + if (p->progress) + { + UInt64 inDelta = p->inProcessed - inPrev; + UInt64 outDelta = p->outProcessed - outPrev; + if (inDelta >= (1 << 22) || outDelta >= (1 << 22)) + { + RINOK(ICompressProgress_Progress(p->progress, p->inProcessed, p->outProcessed)); + inPrev = p->inProcessed; + outPrev = p->outProcessed; + } + } + } +} + + + +SRes Lzma2DecMt_Decode(CLzma2DecMtHandle pp, + Byte prop, + const CLzma2DecMtProps *props, + ISeqOutStream *outStream, const UInt64 *outDataSize, int finishMode, + // Byte *outBuf, size_t *outBufSize, + ISeqInStream *inStream, + // const Byte *inData, size_t inDataSize, + UInt64 *inProcessed, + // UInt64 *outProcessed, + int *isMT, + ICompressProgress *progress) +{ + CLzma2DecMt *p = (CLzma2DecMt *)pp; + #ifndef _7ZIP_ST + BoolInt tMode; + #endif + + *inProcessed = 0; + + if (prop > 40) + return SZ_ERROR_UNSUPPORTED; + + p->prop = prop; + p->props = *props; + + p->inStream = inStream; + p->outStream = outStream; + p->progress = progress; + + p->outSize = 0; + p->outSize_Defined = False; + if (outDataSize) + { + p->outSize_Defined = True; + p->outSize = *outDataSize; + } + p->finishMode = finishMode; + + p->outProcessed = 0; + p->inProcessed = 0; + + p->readWasFinished = False; + p->readRes = SZ_OK; + + *isMT = False; + + + #ifndef _7ZIP_ST + + tMode = False; + + // p->mtc.parseRes = SZ_OK; + + // p->mtc.numFilledThreads = 0; + // p->mtc.crossStart = 0; + // p->mtc.crossEnd = 0; + // p->mtc.allocError_for_Read_BlockIndex = 0; + // p->mtc.isAllocError = False; + + if (p->props.numThreads > 1) + { + IMtDecCallback2 vt; + + Lzma2DecMt_FreeSt(p); + + p->outProcessed_Parse = 0; + + if (!p->mtc_WasConstructed) + { + p->mtc_WasConstructed = True; + MtDec_Construct(&p->mtc); + } + + p->mtc.progress = progress; + p->mtc.inStream = inStream; + + // p->outBuf = NULL; + // p->outBufSize = 0; + /* + if (!outStream) + { + // p->outBuf = outBuf; + // p->outBufSize = *outBufSize; + // *outBufSize = 0; + return SZ_ERROR_PARAM; + } + */ + + // p->mtc.inBlockMax = p->props.inBlockMax; + p->mtc.alloc = &p->alignOffsetAlloc.vt; + // p->alignOffsetAlloc.baseAlloc; + // p->mtc.inData = inData; + // p->mtc.inDataSize = inDataSize; + p->mtc.mtCallback = &vt; + p->mtc.mtCallbackObject = p; + + p->mtc.inBufSize = p->props.inBufSize_MT; + + p->mtc.numThreadsMax = p->props.numThreads; + + *isMT = True; + + vt.Parse = Lzma2DecMt_MtCallback_Parse; + vt.PreCode = Lzma2DecMt_MtCallback_PreCode; + vt.Code = Lzma2DecMt_MtCallback_Code; + vt.Write = Lzma2DecMt_MtCallback_Write; + + { + BoolInt needContinue = False; + + SRes res = MtDec_Code(&p->mtc); + + /* + if (!outStream) + *outBufSize = p->outBuf - outBuf; + */ + + *inProcessed = p->mtc.inProcessed; + + needContinue = False; + + if (res == SZ_OK) + { + if (p->mtc.mtProgress.res != SZ_OK) + res = p->mtc.mtProgress.res; + else + needContinue = p->mtc.needContinue; + } + + if (!needContinue) + { + if (res == SZ_OK) + return p->mtc.readRes; + return res; + } + + tMode = True; + p->readRes = p->mtc.readRes; + p->readWasFinished = p->mtc.readWasFinished; + p->inProcessed = p->mtc.inProcessed; + + PRF_STR("----- decoding ST -----"); + } + } + + #endif + + + *isMT = False; + + { + SRes res = Lzma2Dec_Decode_ST(p + #ifndef _7ZIP_ST + , tMode + #endif + ); + + *inProcessed = p->inProcessed; + + // res = SZ_OK; // for test + if (res == SZ_ERROR_INPUT_EOF) + { + if (p->readRes != SZ_OK) + res = p->readRes; + } + else if (res == SZ_OK && p->readRes != SZ_OK) + res = p->readRes; + + /* + #ifndef _7ZIP_ST + if (res == SZ_OK && tMode && p->mtc.parseRes != SZ_OK) + res = p->mtc.parseRes; + #endif + */ + + return res; + } +} + + +/* ---------- Read from CLzma2DecMtHandle Interface ---------- */ + +SRes Lzma2DecMt_Init(CLzma2DecMtHandle pp, + Byte prop, + const CLzma2DecMtProps *props, + const UInt64 *outDataSize, int finishMode, + ISeqInStream *inStream) +{ + CLzma2DecMt *p = (CLzma2DecMt *)pp; + + if (prop > 40) + return SZ_ERROR_UNSUPPORTED; + + p->prop = prop; + p->props = *props; + + p->inStream = inStream; + + p->outSize = 0; + p->outSize_Defined = False; + if (outDataSize) + { + p->outSize_Defined = True; + p->outSize = *outDataSize; + } + p->finishMode = finishMode; + + p->outProcessed = 0; + p->inProcessed = 0; + + p->inPos = 0; + p->inLim = 0; + + return Lzma2Dec_Prepare_ST(p); +} + + +SRes Lzma2DecMt_Read(CLzma2DecMtHandle pp, + Byte *data, size_t *outSize, + UInt64 *inStreamProcessed) +{ + CLzma2DecMt *p = (CLzma2DecMt *)pp; + ELzmaFinishMode finishMode; + SRes readRes; + size_t size = *outSize; + + *outSize = 0; + *inStreamProcessed = 0; + + finishMode = LZMA_FINISH_ANY; + if (p->outSize_Defined) + { + const UInt64 rem = p->outSize - p->outProcessed; + if (size >= rem) + { + size = (size_t)rem; + if (p->finishMode) + finishMode = LZMA_FINISH_END; + } + } + + readRes = SZ_OK; + + for (;;) + { + SizeT inCur; + SizeT outCur; + ELzmaStatus status; + SRes res; + + if (p->inPos == p->inLim && readRes == SZ_OK) + { + p->inPos = 0; + p->inLim = p->props.inBufSize_ST; + readRes = ISeqInStream_Read(p->inStream, p->inBuf, &p->inLim); + } + + inCur = p->inLim - p->inPos; + outCur = size; + + res = Lzma2Dec_DecodeToBuf(&p->dec, data, &outCur, + p->inBuf + p->inPos, &inCur, finishMode, &status); + + p->inPos += inCur; + p->inProcessed += inCur; + *inStreamProcessed += inCur; + p->outProcessed += outCur; + *outSize += outCur; + size -= outCur; + data += outCur; + + if (res != 0) + return res; + + /* + if (status == LZMA_STATUS_FINISHED_WITH_MARK) + return readRes; + + if (size == 0 && status != LZMA_STATUS_NEEDS_MORE_INPUT) + { + if (p->finishMode && p->outSize_Defined && p->outProcessed >= p->outSize) + return SZ_ERROR_DATA; + return readRes; + } + */ + + if (inCur == 0 && outCur == 0) + return readRes; + } +} diff --git a/deps/LZMA-SDK/C/Lzma2DecMt.h b/deps/LZMA-SDK/C/Lzma2DecMt.h index 96f89a3ca..7791c310b 100644 --- a/deps/LZMA-SDK/C/Lzma2DecMt.h +++ b/deps/LZMA-SDK/C/Lzma2DecMt.h @@ -1,79 +1,79 @@ -/* Lzma2DecMt.h -- LZMA2 Decoder Multi-thread -2018-02-17 : Igor Pavlov : Public domain */ - -#ifndef __LZMA2_DEC_MT_H -#define __LZMA2_DEC_MT_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -typedef struct -{ - size_t inBufSize_ST; - size_t outStep_ST; - - #ifndef _7ZIP_ST - unsigned numThreads; - size_t inBufSize_MT; - size_t outBlockMax; - size_t inBlockMax; - #endif -} CLzma2DecMtProps; - -/* init to single-thread mode */ -void Lzma2DecMtProps_Init(CLzma2DecMtProps *p); - - -/* ---------- CLzma2DecMtHandle Interface ---------- */ - -/* Lzma2DecMt_ * functions can return the following exit codes: -SRes: - SZ_OK - OK - SZ_ERROR_MEM - Memory allocation error - SZ_ERROR_PARAM - Incorrect paramater in props - SZ_ERROR_WRITE - ISeqOutStream write callback error - // SZ_ERROR_OUTPUT_EOF - output buffer overflow - version with (Byte *) output - SZ_ERROR_PROGRESS - some break from progress callback - SZ_ERROR_THREAD - error in multithreading functions (only for Mt version) -*/ - -typedef void * CLzma2DecMtHandle; - -CLzma2DecMtHandle Lzma2DecMt_Create(ISzAllocPtr alloc, ISzAllocPtr allocMid); -void Lzma2DecMt_Destroy(CLzma2DecMtHandle p); - -SRes Lzma2DecMt_Decode(CLzma2DecMtHandle p, - Byte prop, - const CLzma2DecMtProps *props, - ISeqOutStream *outStream, - const UInt64 *outDataSize, // NULL means undefined - int finishMode, // 0 - partial unpacking is allowed, 1 - if lzma2 stream must be finished - // Byte *outBuf, size_t *outBufSize, - ISeqInStream *inStream, - // const Byte *inData, size_t inDataSize, - - // out variables: - UInt64 *inProcessed, - int *isMT, /* out: (*isMT == 0), if single thread decoding was used */ - - // UInt64 *outProcessed, - ICompressProgress *progress); - - -/* ---------- Read from CLzma2DecMtHandle Interface ---------- */ - -SRes Lzma2DecMt_Init(CLzma2DecMtHandle pp, - Byte prop, - const CLzma2DecMtProps *props, - const UInt64 *outDataSize, int finishMode, - ISeqInStream *inStream); - -SRes Lzma2DecMt_Read(CLzma2DecMtHandle pp, - Byte *data, size_t *outSize, - UInt64 *inStreamProcessed); - - -EXTERN_C_END - -#endif +/* Lzma2DecMt.h -- LZMA2 Decoder Multi-thread +2018-02-17 : Igor Pavlov : Public domain */ + +#ifndef __LZMA2_DEC_MT_H +#define __LZMA2_DEC_MT_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +typedef struct +{ + size_t inBufSize_ST; + size_t outStep_ST; + + #ifndef _7ZIP_ST + unsigned numThreads; + size_t inBufSize_MT; + size_t outBlockMax; + size_t inBlockMax; + #endif +} CLzma2DecMtProps; + +/* init to single-thread mode */ +void Lzma2DecMtProps_Init(CLzma2DecMtProps *p); + + +/* ---------- CLzma2DecMtHandle Interface ---------- */ + +/* Lzma2DecMt_ * functions can return the following exit codes: +SRes: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater in props + SZ_ERROR_WRITE - ISeqOutStream write callback error + // SZ_ERROR_OUTPUT_EOF - output buffer overflow - version with (Byte *) output + SZ_ERROR_PROGRESS - some break from progress callback + SZ_ERROR_THREAD - error in multithreading functions (only for Mt version) +*/ + +typedef void * CLzma2DecMtHandle; + +CLzma2DecMtHandle Lzma2DecMt_Create(ISzAllocPtr alloc, ISzAllocPtr allocMid); +void Lzma2DecMt_Destroy(CLzma2DecMtHandle p); + +SRes Lzma2DecMt_Decode(CLzma2DecMtHandle p, + Byte prop, + const CLzma2DecMtProps *props, + ISeqOutStream *outStream, + const UInt64 *outDataSize, // NULL means undefined + int finishMode, // 0 - partial unpacking is allowed, 1 - if lzma2 stream must be finished + // Byte *outBuf, size_t *outBufSize, + ISeqInStream *inStream, + // const Byte *inData, size_t inDataSize, + + // out variables: + UInt64 *inProcessed, + int *isMT, /* out: (*isMT == 0), if single thread decoding was used */ + + // UInt64 *outProcessed, + ICompressProgress *progress); + + +/* ---------- Read from CLzma2DecMtHandle Interface ---------- */ + +SRes Lzma2DecMt_Init(CLzma2DecMtHandle pp, + Byte prop, + const CLzma2DecMtProps *props, + const UInt64 *outDataSize, int finishMode, + ISeqInStream *inStream); + +SRes Lzma2DecMt_Read(CLzma2DecMtHandle pp, + Byte *data, size_t *outSize, + UInt64 *inStreamProcessed); + + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/Lzma2Enc.c b/deps/LZMA-SDK/C/Lzma2Enc.c index c8b114cb4..e61a5dfe0 100644 --- a/deps/LZMA-SDK/C/Lzma2Enc.c +++ b/deps/LZMA-SDK/C/Lzma2Enc.c @@ -1,803 +1,803 @@ -/* Lzma2Enc.c -- LZMA2 Encoder -2021-02-09 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include - -/* #define _7ZIP_ST */ - -#include "Lzma2Enc.h" - -#ifndef _7ZIP_ST -#include "MtCoder.h" -#else -#define MTCODER__THREADS_MAX 1 -#endif - -#define LZMA2_CONTROL_LZMA (1 << 7) -#define LZMA2_CONTROL_COPY_NO_RESET 2 -#define LZMA2_CONTROL_COPY_RESET_DIC 1 -#define LZMA2_CONTROL_EOF 0 - -#define LZMA2_LCLP_MAX 4 - -#define LZMA2_DIC_SIZE_FROM_PROP(p) (((UInt32)2 | ((p) & 1)) << ((p) / 2 + 11)) - -#define LZMA2_PACK_SIZE_MAX (1 << 16) -#define LZMA2_COPY_CHUNK_SIZE LZMA2_PACK_SIZE_MAX -#define LZMA2_UNPACK_SIZE_MAX (1 << 21) -#define LZMA2_KEEP_WINDOW_SIZE LZMA2_UNPACK_SIZE_MAX - -#define LZMA2_CHUNK_SIZE_COMPRESSED_MAX ((1 << 16) + 16) - - -#define PRF(x) /* x */ - - -/* ---------- CLimitedSeqInStream ---------- */ - -typedef struct -{ - ISeqInStream vt; - ISeqInStream *realStream; - UInt64 limit; - UInt64 processed; - int finished; -} CLimitedSeqInStream; - -static void LimitedSeqInStream_Init(CLimitedSeqInStream *p) -{ - p->limit = (UInt64)(Int64)-1; - p->processed = 0; - p->finished = 0; -} - -static SRes LimitedSeqInStream_Read(const ISeqInStream *pp, void *data, size_t *size) -{ - CLimitedSeqInStream *p = CONTAINER_FROM_VTBL(pp, CLimitedSeqInStream, vt); - size_t size2 = *size; - SRes res = SZ_OK; - - if (p->limit != (UInt64)(Int64)-1) - { - UInt64 rem = p->limit - p->processed; - if (size2 > rem) - size2 = (size_t)rem; - } - if (size2 != 0) - { - res = ISeqInStream_Read(p->realStream, data, &size2); - p->finished = (size2 == 0 ? 1 : 0); - p->processed += size2; - } - *size = size2; - return res; -} - - -/* ---------- CLzma2EncInt ---------- */ - -typedef struct -{ - CLzmaEncHandle enc; - Byte propsAreSet; - Byte propsByte; - Byte needInitState; - Byte needInitProp; - UInt64 srcPos; -} CLzma2EncInt; - - -static SRes Lzma2EncInt_InitStream(CLzma2EncInt *p, const CLzma2EncProps *props) -{ - if (!p->propsAreSet) - { - SizeT propsSize = LZMA_PROPS_SIZE; - Byte propsEncoded[LZMA_PROPS_SIZE]; - RINOK(LzmaEnc_SetProps(p->enc, &props->lzmaProps)); - RINOK(LzmaEnc_WriteProperties(p->enc, propsEncoded, &propsSize)); - p->propsByte = propsEncoded[0]; - p->propsAreSet = True; - } - return SZ_OK; -} - -static void Lzma2EncInt_InitBlock(CLzma2EncInt *p) -{ - p->srcPos = 0; - p->needInitState = True; - p->needInitProp = True; -} - - -SRes LzmaEnc_PrepareForLzma2(CLzmaEncHandle pp, ISeqInStream *inStream, UInt32 keepWindowSize, - ISzAllocPtr alloc, ISzAllocPtr allocBig); -SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen, - UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig); -SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit, - Byte *dest, size_t *destLen, UInt32 desiredPackSize, UInt32 *unpackSize); -const Byte *LzmaEnc_GetCurBuf(CLzmaEncHandle pp); -void LzmaEnc_Finish(CLzmaEncHandle pp); -void LzmaEnc_SaveState(CLzmaEncHandle pp); -void LzmaEnc_RestoreState(CLzmaEncHandle pp); - -/* -UInt32 LzmaEnc_GetNumAvailableBytes(CLzmaEncHandle pp); -*/ - -static SRes Lzma2EncInt_EncodeSubblock(CLzma2EncInt *p, Byte *outBuf, - size_t *packSizeRes, ISeqOutStream *outStream) -{ - size_t packSizeLimit = *packSizeRes; - size_t packSize = packSizeLimit; - UInt32 unpackSize = LZMA2_UNPACK_SIZE_MAX; - unsigned lzHeaderSize = 5 + (p->needInitProp ? 1 : 0); - BoolInt useCopyBlock; - SRes res; - - *packSizeRes = 0; - if (packSize < lzHeaderSize) - return SZ_ERROR_OUTPUT_EOF; - packSize -= lzHeaderSize; - - LzmaEnc_SaveState(p->enc); - res = LzmaEnc_CodeOneMemBlock(p->enc, p->needInitState, - outBuf + lzHeaderSize, &packSize, LZMA2_PACK_SIZE_MAX, &unpackSize); - - PRF(printf("\npackSize = %7d unpackSize = %7d ", packSize, unpackSize)); - - if (unpackSize == 0) - return res; - - if (res == SZ_OK) - useCopyBlock = (packSize + 2 >= unpackSize || packSize > (1 << 16)); - else - { - if (res != SZ_ERROR_OUTPUT_EOF) - return res; - res = SZ_OK; - useCopyBlock = True; - } - - if (useCopyBlock) - { - size_t destPos = 0; - PRF(printf("################# COPY ")); - - while (unpackSize > 0) - { - UInt32 u = (unpackSize < LZMA2_COPY_CHUNK_SIZE) ? unpackSize : LZMA2_COPY_CHUNK_SIZE; - if (packSizeLimit - destPos < u + 3) - return SZ_ERROR_OUTPUT_EOF; - outBuf[destPos++] = (Byte)(p->srcPos == 0 ? LZMA2_CONTROL_COPY_RESET_DIC : LZMA2_CONTROL_COPY_NO_RESET); - outBuf[destPos++] = (Byte)((u - 1) >> 8); - outBuf[destPos++] = (Byte)(u - 1); - memcpy(outBuf + destPos, LzmaEnc_GetCurBuf(p->enc) - unpackSize, u); - unpackSize -= u; - destPos += u; - p->srcPos += u; - - if (outStream) - { - *packSizeRes += destPos; - if (ISeqOutStream_Write(outStream, outBuf, destPos) != destPos) - return SZ_ERROR_WRITE; - destPos = 0; - } - else - *packSizeRes = destPos; - /* needInitState = True; */ - } - - LzmaEnc_RestoreState(p->enc); - return SZ_OK; - } - - { - size_t destPos = 0; - UInt32 u = unpackSize - 1; - UInt32 pm = (UInt32)(packSize - 1); - unsigned mode = (p->srcPos == 0) ? 3 : (p->needInitState ? (p->needInitProp ? 2 : 1) : 0); - - PRF(printf(" ")); - - outBuf[destPos++] = (Byte)(LZMA2_CONTROL_LZMA | (mode << 5) | ((u >> 16) & 0x1F)); - outBuf[destPos++] = (Byte)(u >> 8); - outBuf[destPos++] = (Byte)u; - outBuf[destPos++] = (Byte)(pm >> 8); - outBuf[destPos++] = (Byte)pm; - - if (p->needInitProp) - outBuf[destPos++] = p->propsByte; - - p->needInitProp = False; - p->needInitState = False; - destPos += packSize; - p->srcPos += unpackSize; - - if (outStream) - if (ISeqOutStream_Write(outStream, outBuf, destPos) != destPos) - return SZ_ERROR_WRITE; - - *packSizeRes = destPos; - return SZ_OK; - } -} - - -/* ---------- Lzma2 Props ---------- */ - -void Lzma2EncProps_Init(CLzma2EncProps *p) -{ - LzmaEncProps_Init(&p->lzmaProps); - p->blockSize = LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO; - p->numBlockThreads_Reduced = -1; - p->numBlockThreads_Max = -1; - p->numTotalThreads = -1; -} - -void Lzma2EncProps_Normalize(CLzma2EncProps *p) -{ - UInt64 fileSize; - int t1, t1n, t2, t2r, t3; - { - CLzmaEncProps lzmaProps = p->lzmaProps; - LzmaEncProps_Normalize(&lzmaProps); - t1n = lzmaProps.numThreads; - } - - t1 = p->lzmaProps.numThreads; - t2 = p->numBlockThreads_Max; - t3 = p->numTotalThreads; - - if (t2 > MTCODER__THREADS_MAX) - t2 = MTCODER__THREADS_MAX; - - if (t3 <= 0) - { - if (t2 <= 0) - t2 = 1; - t3 = t1n * t2; - } - else if (t2 <= 0) - { - t2 = t3 / t1n; - if (t2 == 0) - { - t1 = 1; - t2 = t3; - } - if (t2 > MTCODER__THREADS_MAX) - t2 = MTCODER__THREADS_MAX; - } - else if (t1 <= 0) - { - t1 = t3 / t2; - if (t1 == 0) - t1 = 1; - } - else - t3 = t1n * t2; - - p->lzmaProps.numThreads = t1; - - t2r = t2; - - fileSize = p->lzmaProps.reduceSize; - - if ( p->blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID - && p->blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO - && (p->blockSize < fileSize || fileSize == (UInt64)(Int64)-1)) - p->lzmaProps.reduceSize = p->blockSize; - - LzmaEncProps_Normalize(&p->lzmaProps); - - p->lzmaProps.reduceSize = fileSize; - - t1 = p->lzmaProps.numThreads; - - if (p->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID) - { - t2r = t2 = 1; - t3 = t1; - } - else if (p->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO && t2 <= 1) - { - /* if there is no block multi-threading, we use SOLID block */ - p->blockSize = LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID; - } - else - { - if (p->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO) - { - const UInt32 kMinSize = (UInt32)1 << 20; - const UInt32 kMaxSize = (UInt32)1 << 28; - const UInt32 dictSize = p->lzmaProps.dictSize; - UInt64 blockSize = (UInt64)dictSize << 2; - if (blockSize < kMinSize) blockSize = kMinSize; - if (blockSize > kMaxSize) blockSize = kMaxSize; - if (blockSize < dictSize) blockSize = dictSize; - blockSize += (kMinSize - 1); - blockSize &= ~(UInt64)(kMinSize - 1); - p->blockSize = blockSize; - } - - if (t2 > 1 && fileSize != (UInt64)(Int64)-1) - { - UInt64 numBlocks = fileSize / p->blockSize; - if (numBlocks * p->blockSize != fileSize) - numBlocks++; - if (numBlocks < (unsigned)t2) - { - t2r = (int)numBlocks; - if (t2r == 0) - t2r = 1; - t3 = t1 * t2r; - } - } - } - - p->numBlockThreads_Max = t2; - p->numBlockThreads_Reduced = t2r; - p->numTotalThreads = t3; -} - - -static SRes Progress(ICompressProgress *p, UInt64 inSize, UInt64 outSize) -{ - return (p && ICompressProgress_Progress(p, inSize, outSize) != SZ_OK) ? SZ_ERROR_PROGRESS : SZ_OK; -} - - -/* ---------- Lzma2 ---------- */ - -typedef struct -{ - Byte propEncoded; - CLzma2EncProps props; - UInt64 expectedDataSize; - - Byte *tempBufLzma; - - ISzAllocPtr alloc; - ISzAllocPtr allocBig; - - CLzma2EncInt coders[MTCODER__THREADS_MAX]; - - #ifndef _7ZIP_ST - - ISeqOutStream *outStream; - Byte *outBuf; - size_t outBuf_Rem; /* remainder in outBuf */ - - size_t outBufSize; /* size of allocated outBufs[i] */ - size_t outBufsDataSizes[MTCODER__BLOCKS_MAX]; - BoolInt mtCoder_WasConstructed; - CMtCoder mtCoder; - Byte *outBufs[MTCODER__BLOCKS_MAX]; - - #endif - -} CLzma2Enc; - - - -CLzma2EncHandle Lzma2Enc_Create(ISzAllocPtr alloc, ISzAllocPtr allocBig) -{ - CLzma2Enc *p = (CLzma2Enc *)ISzAlloc_Alloc(alloc, sizeof(CLzma2Enc)); - if (!p) - return NULL; - Lzma2EncProps_Init(&p->props); - Lzma2EncProps_Normalize(&p->props); - p->expectedDataSize = (UInt64)(Int64)-1; - p->tempBufLzma = NULL; - p->alloc = alloc; - p->allocBig = allocBig; - { - unsigned i; - for (i = 0; i < MTCODER__THREADS_MAX; i++) - p->coders[i].enc = NULL; - } - - #ifndef _7ZIP_ST - p->mtCoder_WasConstructed = False; - { - unsigned i; - for (i = 0; i < MTCODER__BLOCKS_MAX; i++) - p->outBufs[i] = NULL; - p->outBufSize = 0; - } - #endif - - return p; -} - - -#ifndef _7ZIP_ST - -static void Lzma2Enc_FreeOutBufs(CLzma2Enc *p) -{ - unsigned i; - for (i = 0; i < MTCODER__BLOCKS_MAX; i++) - if (p->outBufs[i]) - { - ISzAlloc_Free(p->alloc, p->outBufs[i]); - p->outBufs[i] = NULL; - } - p->outBufSize = 0; -} - -#endif - - -void Lzma2Enc_Destroy(CLzma2EncHandle pp) -{ - CLzma2Enc *p = (CLzma2Enc *)pp; - unsigned i; - for (i = 0; i < MTCODER__THREADS_MAX; i++) - { - CLzma2EncInt *t = &p->coders[i]; - if (t->enc) - { - LzmaEnc_Destroy(t->enc, p->alloc, p->allocBig); - t->enc = NULL; - } - } - - - #ifndef _7ZIP_ST - if (p->mtCoder_WasConstructed) - { - MtCoder_Destruct(&p->mtCoder); - p->mtCoder_WasConstructed = False; - } - Lzma2Enc_FreeOutBufs(p); - #endif - - ISzAlloc_Free(p->alloc, p->tempBufLzma); - p->tempBufLzma = NULL; - - ISzAlloc_Free(p->alloc, pp); -} - - -SRes Lzma2Enc_SetProps(CLzma2EncHandle pp, const CLzma2EncProps *props) -{ - CLzma2Enc *p = (CLzma2Enc *)pp; - CLzmaEncProps lzmaProps = props->lzmaProps; - LzmaEncProps_Normalize(&lzmaProps); - if (lzmaProps.lc + lzmaProps.lp > LZMA2_LCLP_MAX) - return SZ_ERROR_PARAM; - p->props = *props; - Lzma2EncProps_Normalize(&p->props); - return SZ_OK; -} - - -void Lzma2Enc_SetDataSize(CLzmaEncHandle pp, UInt64 expectedDataSiize) -{ - CLzma2Enc *p = (CLzma2Enc *)pp; - p->expectedDataSize = expectedDataSiize; -} - - -Byte Lzma2Enc_WriteProperties(CLzma2EncHandle pp) -{ - CLzma2Enc *p = (CLzma2Enc *)pp; - unsigned i; - UInt32 dicSize = LzmaEncProps_GetDictSize(&p->props.lzmaProps); - for (i = 0; i < 40; i++) - if (dicSize <= LZMA2_DIC_SIZE_FROM_PROP(i)) - break; - return (Byte)i; -} - - -static SRes Lzma2Enc_EncodeMt1( - CLzma2Enc *me, - CLzma2EncInt *p, - ISeqOutStream *outStream, - Byte *outBuf, size_t *outBufSize, - ISeqInStream *inStream, - const Byte *inData, size_t inDataSize, - int finished, - ICompressProgress *progress) -{ - UInt64 unpackTotal = 0; - UInt64 packTotal = 0; - size_t outLim = 0; - CLimitedSeqInStream limitedInStream; - - if (outBuf) - { - outLim = *outBufSize; - *outBufSize = 0; - } - - if (!p->enc) - { - p->propsAreSet = False; - p->enc = LzmaEnc_Create(me->alloc); - if (!p->enc) - return SZ_ERROR_MEM; - } - - limitedInStream.realStream = inStream; - if (inStream) - { - limitedInStream.vt.Read = LimitedSeqInStream_Read; - } - - if (!outBuf) - { - // outStream version works only in one thread. So we use CLzma2Enc::tempBufLzma - if (!me->tempBufLzma) - { - me->tempBufLzma = (Byte *)ISzAlloc_Alloc(me->alloc, LZMA2_CHUNK_SIZE_COMPRESSED_MAX); - if (!me->tempBufLzma) - return SZ_ERROR_MEM; - } - } - - RINOK(Lzma2EncInt_InitStream(p, &me->props)); - - for (;;) - { - SRes res = SZ_OK; - size_t inSizeCur = 0; - - Lzma2EncInt_InitBlock(p); - - LimitedSeqInStream_Init(&limitedInStream); - limitedInStream.limit = me->props.blockSize; - - if (inStream) - { - UInt64 expected = (UInt64)(Int64)-1; - // inStream version works only in one thread. So we use CLzma2Enc::expectedDataSize - if (me->expectedDataSize != (UInt64)(Int64)-1 - && me->expectedDataSize >= unpackTotal) - expected = me->expectedDataSize - unpackTotal; - if (me->props.blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID - && expected > me->props.blockSize) - expected = (size_t)me->props.blockSize; - - LzmaEnc_SetDataSize(p->enc, expected); - - RINOK(LzmaEnc_PrepareForLzma2(p->enc, - &limitedInStream.vt, - LZMA2_KEEP_WINDOW_SIZE, - me->alloc, - me->allocBig)); - } - else - { - inSizeCur = inDataSize - (size_t)unpackTotal; - if (me->props.blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID - && inSizeCur > me->props.blockSize) - inSizeCur = (size_t)me->props.blockSize; - - // LzmaEnc_SetDataSize(p->enc, inSizeCur); - - RINOK(LzmaEnc_MemPrepare(p->enc, - inData + (size_t)unpackTotal, inSizeCur, - LZMA2_KEEP_WINDOW_SIZE, - me->alloc, - me->allocBig)); - } - - for (;;) - { - size_t packSize = LZMA2_CHUNK_SIZE_COMPRESSED_MAX; - if (outBuf) - packSize = outLim - (size_t)packTotal; - - res = Lzma2EncInt_EncodeSubblock(p, - outBuf ? outBuf + (size_t)packTotal : me->tempBufLzma, &packSize, - outBuf ? NULL : outStream); - - if (res != SZ_OK) - break; - - packTotal += packSize; - if (outBuf) - *outBufSize = (size_t)packTotal; - - res = Progress(progress, unpackTotal + p->srcPos, packTotal); - if (res != SZ_OK) - break; - - /* - if (LzmaEnc_GetNumAvailableBytes(p->enc) == 0) - break; - */ - - if (packSize == 0) - break; - } - - LzmaEnc_Finish(p->enc); - - unpackTotal += p->srcPos; - - RINOK(res); - - if (p->srcPos != (inStream ? limitedInStream.processed : inSizeCur)) - return SZ_ERROR_FAIL; - - if (inStream ? limitedInStream.finished : (unpackTotal == inDataSize)) - { - if (finished) - { - if (outBuf) - { - const size_t destPos = *outBufSize; - if (destPos >= outLim) - return SZ_ERROR_OUTPUT_EOF; - outBuf[destPos] = LZMA2_CONTROL_EOF; // 0 - *outBufSize = destPos + 1; - } - else - { - const Byte b = LZMA2_CONTROL_EOF; // 0; - if (ISeqOutStream_Write(outStream, &b, 1) != 1) - return SZ_ERROR_WRITE; - } - } - return SZ_OK; - } - } -} - - - -#ifndef _7ZIP_ST - -static SRes Lzma2Enc_MtCallback_Code(void *pp, unsigned coderIndex, unsigned outBufIndex, - const Byte *src, size_t srcSize, int finished) -{ - CLzma2Enc *me = (CLzma2Enc *)pp; - size_t destSize = me->outBufSize; - SRes res; - CMtProgressThunk progressThunk; - - Byte *dest = me->outBufs[outBufIndex]; - - me->outBufsDataSizes[outBufIndex] = 0; - - if (!dest) - { - dest = (Byte *)ISzAlloc_Alloc(me->alloc, me->outBufSize); - if (!dest) - return SZ_ERROR_MEM; - me->outBufs[outBufIndex] = dest; - } - - MtProgressThunk_CreateVTable(&progressThunk); - progressThunk.mtProgress = &me->mtCoder.mtProgress; - progressThunk.inSize = 0; - progressThunk.outSize = 0; - - res = Lzma2Enc_EncodeMt1(me, - &me->coders[coderIndex], - NULL, dest, &destSize, - NULL, src, srcSize, - finished, - &progressThunk.vt); - - me->outBufsDataSizes[outBufIndex] = destSize; - - return res; -} - - -static SRes Lzma2Enc_MtCallback_Write(void *pp, unsigned outBufIndex) -{ - CLzma2Enc *me = (CLzma2Enc *)pp; - size_t size = me->outBufsDataSizes[outBufIndex]; - const Byte *data = me->outBufs[outBufIndex]; - - if (me->outStream) - return ISeqOutStream_Write(me->outStream, data, size) == size ? SZ_OK : SZ_ERROR_WRITE; - - if (size > me->outBuf_Rem) - return SZ_ERROR_OUTPUT_EOF; - memcpy(me->outBuf, data, size); - me->outBuf_Rem -= size; - me->outBuf += size; - return SZ_OK; -} - -#endif - - - -SRes Lzma2Enc_Encode2(CLzma2EncHandle pp, - ISeqOutStream *outStream, - Byte *outBuf, size_t *outBufSize, - ISeqInStream *inStream, - const Byte *inData, size_t inDataSize, - ICompressProgress *progress) -{ - CLzma2Enc *p = (CLzma2Enc *)pp; - - if (inStream && inData) - return SZ_ERROR_PARAM; - - if (outStream && outBuf) - return SZ_ERROR_PARAM; - - { - unsigned i; - for (i = 0; i < MTCODER__THREADS_MAX; i++) - p->coders[i].propsAreSet = False; - } - - #ifndef _7ZIP_ST - - if (p->props.numBlockThreads_Reduced > 1) - { - IMtCoderCallback2 vt; - - if (!p->mtCoder_WasConstructed) - { - p->mtCoder_WasConstructed = True; - MtCoder_Construct(&p->mtCoder); - } - - vt.Code = Lzma2Enc_MtCallback_Code; - vt.Write = Lzma2Enc_MtCallback_Write; - - p->outStream = outStream; - p->outBuf = NULL; - p->outBuf_Rem = 0; - if (!outStream) - { - p->outBuf = outBuf; - p->outBuf_Rem = *outBufSize; - *outBufSize = 0; - } - - p->mtCoder.allocBig = p->allocBig; - p->mtCoder.progress = progress; - p->mtCoder.inStream = inStream; - p->mtCoder.inData = inData; - p->mtCoder.inDataSize = inDataSize; - p->mtCoder.mtCallback = &vt; - p->mtCoder.mtCallbackObject = p; - - p->mtCoder.blockSize = (size_t)p->props.blockSize; - if (p->mtCoder.blockSize != p->props.blockSize) - return SZ_ERROR_PARAM; /* SZ_ERROR_MEM */ - - { - size_t destBlockSize = p->mtCoder.blockSize + (p->mtCoder.blockSize >> 10) + 16; - if (destBlockSize < p->mtCoder.blockSize) - return SZ_ERROR_PARAM; - if (p->outBufSize != destBlockSize) - Lzma2Enc_FreeOutBufs(p); - p->outBufSize = destBlockSize; - } - - p->mtCoder.numThreadsMax = (unsigned)p->props.numBlockThreads_Max; - p->mtCoder.expectedDataSize = p->expectedDataSize; - - { - SRes res = MtCoder_Code(&p->mtCoder); - if (!outStream) - *outBufSize = (size_t)(p->outBuf - outBuf); - return res; - } - } - - #endif - - - return Lzma2Enc_EncodeMt1(p, - &p->coders[0], - outStream, outBuf, outBufSize, - inStream, inData, inDataSize, - True, /* finished */ - progress); -} +/* Lzma2Enc.c -- LZMA2 Encoder +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +/* #define _7ZIP_ST */ + +#include "Lzma2Enc.h" + +#ifndef _7ZIP_ST +#include "MtCoder.h" +#else +#define MTCODER__THREADS_MAX 1 +#endif + +#define LZMA2_CONTROL_LZMA (1 << 7) +#define LZMA2_CONTROL_COPY_NO_RESET 2 +#define LZMA2_CONTROL_COPY_RESET_DIC 1 +#define LZMA2_CONTROL_EOF 0 + +#define LZMA2_LCLP_MAX 4 + +#define LZMA2_DIC_SIZE_FROM_PROP(p) (((UInt32)2 | ((p) & 1)) << ((p) / 2 + 11)) + +#define LZMA2_PACK_SIZE_MAX (1 << 16) +#define LZMA2_COPY_CHUNK_SIZE LZMA2_PACK_SIZE_MAX +#define LZMA2_UNPACK_SIZE_MAX (1 << 21) +#define LZMA2_KEEP_WINDOW_SIZE LZMA2_UNPACK_SIZE_MAX + +#define LZMA2_CHUNK_SIZE_COMPRESSED_MAX ((1 << 16) + 16) + + +#define PRF(x) /* x */ + + +/* ---------- CLimitedSeqInStream ---------- */ + +typedef struct +{ + ISeqInStream vt; + ISeqInStream *realStream; + UInt64 limit; + UInt64 processed; + int finished; +} CLimitedSeqInStream; + +static void LimitedSeqInStream_Init(CLimitedSeqInStream *p) +{ + p->limit = (UInt64)(Int64)-1; + p->processed = 0; + p->finished = 0; +} + +static SRes LimitedSeqInStream_Read(const ISeqInStream *pp, void *data, size_t *size) +{ + CLimitedSeqInStream *p = CONTAINER_FROM_VTBL(pp, CLimitedSeqInStream, vt); + size_t size2 = *size; + SRes res = SZ_OK; + + if (p->limit != (UInt64)(Int64)-1) + { + UInt64 rem = p->limit - p->processed; + if (size2 > rem) + size2 = (size_t)rem; + } + if (size2 != 0) + { + res = ISeqInStream_Read(p->realStream, data, &size2); + p->finished = (size2 == 0 ? 1 : 0); + p->processed += size2; + } + *size = size2; + return res; +} + + +/* ---------- CLzma2EncInt ---------- */ + +typedef struct +{ + CLzmaEncHandle enc; + Byte propsAreSet; + Byte propsByte; + Byte needInitState; + Byte needInitProp; + UInt64 srcPos; +} CLzma2EncInt; + + +static SRes Lzma2EncInt_InitStream(CLzma2EncInt *p, const CLzma2EncProps *props) +{ + if (!p->propsAreSet) + { + SizeT propsSize = LZMA_PROPS_SIZE; + Byte propsEncoded[LZMA_PROPS_SIZE]; + RINOK(LzmaEnc_SetProps(p->enc, &props->lzmaProps)); + RINOK(LzmaEnc_WriteProperties(p->enc, propsEncoded, &propsSize)); + p->propsByte = propsEncoded[0]; + p->propsAreSet = True; + } + return SZ_OK; +} + +static void Lzma2EncInt_InitBlock(CLzma2EncInt *p) +{ + p->srcPos = 0; + p->needInitState = True; + p->needInitProp = True; +} + + +SRes LzmaEnc_PrepareForLzma2(CLzmaEncHandle pp, ISeqInStream *inStream, UInt32 keepWindowSize, + ISzAllocPtr alloc, ISzAllocPtr allocBig); +SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen, + UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig); +SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit, + Byte *dest, size_t *destLen, UInt32 desiredPackSize, UInt32 *unpackSize); +const Byte *LzmaEnc_GetCurBuf(CLzmaEncHandle pp); +void LzmaEnc_Finish(CLzmaEncHandle pp); +void LzmaEnc_SaveState(CLzmaEncHandle pp); +void LzmaEnc_RestoreState(CLzmaEncHandle pp); + +/* +UInt32 LzmaEnc_GetNumAvailableBytes(CLzmaEncHandle pp); +*/ + +static SRes Lzma2EncInt_EncodeSubblock(CLzma2EncInt *p, Byte *outBuf, + size_t *packSizeRes, ISeqOutStream *outStream) +{ + size_t packSizeLimit = *packSizeRes; + size_t packSize = packSizeLimit; + UInt32 unpackSize = LZMA2_UNPACK_SIZE_MAX; + unsigned lzHeaderSize = 5 + (p->needInitProp ? 1 : 0); + BoolInt useCopyBlock; + SRes res; + + *packSizeRes = 0; + if (packSize < lzHeaderSize) + return SZ_ERROR_OUTPUT_EOF; + packSize -= lzHeaderSize; + + LzmaEnc_SaveState(p->enc); + res = LzmaEnc_CodeOneMemBlock(p->enc, p->needInitState, + outBuf + lzHeaderSize, &packSize, LZMA2_PACK_SIZE_MAX, &unpackSize); + + PRF(printf("\npackSize = %7d unpackSize = %7d ", packSize, unpackSize)); + + if (unpackSize == 0) + return res; + + if (res == SZ_OK) + useCopyBlock = (packSize + 2 >= unpackSize || packSize > (1 << 16)); + else + { + if (res != SZ_ERROR_OUTPUT_EOF) + return res; + res = SZ_OK; + useCopyBlock = True; + } + + if (useCopyBlock) + { + size_t destPos = 0; + PRF(printf("################# COPY ")); + + while (unpackSize > 0) + { + UInt32 u = (unpackSize < LZMA2_COPY_CHUNK_SIZE) ? unpackSize : LZMA2_COPY_CHUNK_SIZE; + if (packSizeLimit - destPos < u + 3) + return SZ_ERROR_OUTPUT_EOF; + outBuf[destPos++] = (Byte)(p->srcPos == 0 ? LZMA2_CONTROL_COPY_RESET_DIC : LZMA2_CONTROL_COPY_NO_RESET); + outBuf[destPos++] = (Byte)((u - 1) >> 8); + outBuf[destPos++] = (Byte)(u - 1); + memcpy(outBuf + destPos, LzmaEnc_GetCurBuf(p->enc) - unpackSize, u); + unpackSize -= u; + destPos += u; + p->srcPos += u; + + if (outStream) + { + *packSizeRes += destPos; + if (ISeqOutStream_Write(outStream, outBuf, destPos) != destPos) + return SZ_ERROR_WRITE; + destPos = 0; + } + else + *packSizeRes = destPos; + /* needInitState = True; */ + } + + LzmaEnc_RestoreState(p->enc); + return SZ_OK; + } + + { + size_t destPos = 0; + UInt32 u = unpackSize - 1; + UInt32 pm = (UInt32)(packSize - 1); + unsigned mode = (p->srcPos == 0) ? 3 : (p->needInitState ? (p->needInitProp ? 2 : 1) : 0); + + PRF(printf(" ")); + + outBuf[destPos++] = (Byte)(LZMA2_CONTROL_LZMA | (mode << 5) | ((u >> 16) & 0x1F)); + outBuf[destPos++] = (Byte)(u >> 8); + outBuf[destPos++] = (Byte)u; + outBuf[destPos++] = (Byte)(pm >> 8); + outBuf[destPos++] = (Byte)pm; + + if (p->needInitProp) + outBuf[destPos++] = p->propsByte; + + p->needInitProp = False; + p->needInitState = False; + destPos += packSize; + p->srcPos += unpackSize; + + if (outStream) + if (ISeqOutStream_Write(outStream, outBuf, destPos) != destPos) + return SZ_ERROR_WRITE; + + *packSizeRes = destPos; + return SZ_OK; + } +} + + +/* ---------- Lzma2 Props ---------- */ + +void Lzma2EncProps_Init(CLzma2EncProps *p) +{ + LzmaEncProps_Init(&p->lzmaProps); + p->blockSize = LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO; + p->numBlockThreads_Reduced = -1; + p->numBlockThreads_Max = -1; + p->numTotalThreads = -1; +} + +void Lzma2EncProps_Normalize(CLzma2EncProps *p) +{ + UInt64 fileSize; + int t1, t1n, t2, t2r, t3; + { + CLzmaEncProps lzmaProps = p->lzmaProps; + LzmaEncProps_Normalize(&lzmaProps); + t1n = lzmaProps.numThreads; + } + + t1 = p->lzmaProps.numThreads; + t2 = p->numBlockThreads_Max; + t3 = p->numTotalThreads; + + if (t2 > MTCODER__THREADS_MAX) + t2 = MTCODER__THREADS_MAX; + + if (t3 <= 0) + { + if (t2 <= 0) + t2 = 1; + t3 = t1n * t2; + } + else if (t2 <= 0) + { + t2 = t3 / t1n; + if (t2 == 0) + { + t1 = 1; + t2 = t3; + } + if (t2 > MTCODER__THREADS_MAX) + t2 = MTCODER__THREADS_MAX; + } + else if (t1 <= 0) + { + t1 = t3 / t2; + if (t1 == 0) + t1 = 1; + } + else + t3 = t1n * t2; + + p->lzmaProps.numThreads = t1; + + t2r = t2; + + fileSize = p->lzmaProps.reduceSize; + + if ( p->blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID + && p->blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO + && (p->blockSize < fileSize || fileSize == (UInt64)(Int64)-1)) + p->lzmaProps.reduceSize = p->blockSize; + + LzmaEncProps_Normalize(&p->lzmaProps); + + p->lzmaProps.reduceSize = fileSize; + + t1 = p->lzmaProps.numThreads; + + if (p->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID) + { + t2r = t2 = 1; + t3 = t1; + } + else if (p->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO && t2 <= 1) + { + /* if there is no block multi-threading, we use SOLID block */ + p->blockSize = LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID; + } + else + { + if (p->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO) + { + const UInt32 kMinSize = (UInt32)1 << 20; + const UInt32 kMaxSize = (UInt32)1 << 28; + const UInt32 dictSize = p->lzmaProps.dictSize; + UInt64 blockSize = (UInt64)dictSize << 2; + if (blockSize < kMinSize) blockSize = kMinSize; + if (blockSize > kMaxSize) blockSize = kMaxSize; + if (blockSize < dictSize) blockSize = dictSize; + blockSize += (kMinSize - 1); + blockSize &= ~(UInt64)(kMinSize - 1); + p->blockSize = blockSize; + } + + if (t2 > 1 && fileSize != (UInt64)(Int64)-1) + { + UInt64 numBlocks = fileSize / p->blockSize; + if (numBlocks * p->blockSize != fileSize) + numBlocks++; + if (numBlocks < (unsigned)t2) + { + t2r = (int)numBlocks; + if (t2r == 0) + t2r = 1; + t3 = t1 * t2r; + } + } + } + + p->numBlockThreads_Max = t2; + p->numBlockThreads_Reduced = t2r; + p->numTotalThreads = t3; +} + + +static SRes Progress(ICompressProgress *p, UInt64 inSize, UInt64 outSize) +{ + return (p && ICompressProgress_Progress(p, inSize, outSize) != SZ_OK) ? SZ_ERROR_PROGRESS : SZ_OK; +} + + +/* ---------- Lzma2 ---------- */ + +typedef struct +{ + Byte propEncoded; + CLzma2EncProps props; + UInt64 expectedDataSize; + + Byte *tempBufLzma; + + ISzAllocPtr alloc; + ISzAllocPtr allocBig; + + CLzma2EncInt coders[MTCODER__THREADS_MAX]; + + #ifndef _7ZIP_ST + + ISeqOutStream *outStream; + Byte *outBuf; + size_t outBuf_Rem; /* remainder in outBuf */ + + size_t outBufSize; /* size of allocated outBufs[i] */ + size_t outBufsDataSizes[MTCODER__BLOCKS_MAX]; + BoolInt mtCoder_WasConstructed; + CMtCoder mtCoder; + Byte *outBufs[MTCODER__BLOCKS_MAX]; + + #endif + +} CLzma2Enc; + + + +CLzma2EncHandle Lzma2Enc_Create(ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CLzma2Enc *p = (CLzma2Enc *)ISzAlloc_Alloc(alloc, sizeof(CLzma2Enc)); + if (!p) + return NULL; + Lzma2EncProps_Init(&p->props); + Lzma2EncProps_Normalize(&p->props); + p->expectedDataSize = (UInt64)(Int64)-1; + p->tempBufLzma = NULL; + p->alloc = alloc; + p->allocBig = allocBig; + { + unsigned i; + for (i = 0; i < MTCODER__THREADS_MAX; i++) + p->coders[i].enc = NULL; + } + + #ifndef _7ZIP_ST + p->mtCoder_WasConstructed = False; + { + unsigned i; + for (i = 0; i < MTCODER__BLOCKS_MAX; i++) + p->outBufs[i] = NULL; + p->outBufSize = 0; + } + #endif + + return p; +} + + +#ifndef _7ZIP_ST + +static void Lzma2Enc_FreeOutBufs(CLzma2Enc *p) +{ + unsigned i; + for (i = 0; i < MTCODER__BLOCKS_MAX; i++) + if (p->outBufs[i]) + { + ISzAlloc_Free(p->alloc, p->outBufs[i]); + p->outBufs[i] = NULL; + } + p->outBufSize = 0; +} + +#endif + + +void Lzma2Enc_Destroy(CLzma2EncHandle pp) +{ + CLzma2Enc *p = (CLzma2Enc *)pp; + unsigned i; + for (i = 0; i < MTCODER__THREADS_MAX; i++) + { + CLzma2EncInt *t = &p->coders[i]; + if (t->enc) + { + LzmaEnc_Destroy(t->enc, p->alloc, p->allocBig); + t->enc = NULL; + } + } + + + #ifndef _7ZIP_ST + if (p->mtCoder_WasConstructed) + { + MtCoder_Destruct(&p->mtCoder); + p->mtCoder_WasConstructed = False; + } + Lzma2Enc_FreeOutBufs(p); + #endif + + ISzAlloc_Free(p->alloc, p->tempBufLzma); + p->tempBufLzma = NULL; + + ISzAlloc_Free(p->alloc, pp); +} + + +SRes Lzma2Enc_SetProps(CLzma2EncHandle pp, const CLzma2EncProps *props) +{ + CLzma2Enc *p = (CLzma2Enc *)pp; + CLzmaEncProps lzmaProps = props->lzmaProps; + LzmaEncProps_Normalize(&lzmaProps); + if (lzmaProps.lc + lzmaProps.lp > LZMA2_LCLP_MAX) + return SZ_ERROR_PARAM; + p->props = *props; + Lzma2EncProps_Normalize(&p->props); + return SZ_OK; +} + + +void Lzma2Enc_SetDataSize(CLzmaEncHandle pp, UInt64 expectedDataSiize) +{ + CLzma2Enc *p = (CLzma2Enc *)pp; + p->expectedDataSize = expectedDataSiize; +} + + +Byte Lzma2Enc_WriteProperties(CLzma2EncHandle pp) +{ + CLzma2Enc *p = (CLzma2Enc *)pp; + unsigned i; + UInt32 dicSize = LzmaEncProps_GetDictSize(&p->props.lzmaProps); + for (i = 0; i < 40; i++) + if (dicSize <= LZMA2_DIC_SIZE_FROM_PROP(i)) + break; + return (Byte)i; +} + + +static SRes Lzma2Enc_EncodeMt1( + CLzma2Enc *me, + CLzma2EncInt *p, + ISeqOutStream *outStream, + Byte *outBuf, size_t *outBufSize, + ISeqInStream *inStream, + const Byte *inData, size_t inDataSize, + int finished, + ICompressProgress *progress) +{ + UInt64 unpackTotal = 0; + UInt64 packTotal = 0; + size_t outLim = 0; + CLimitedSeqInStream limitedInStream; + + if (outBuf) + { + outLim = *outBufSize; + *outBufSize = 0; + } + + if (!p->enc) + { + p->propsAreSet = False; + p->enc = LzmaEnc_Create(me->alloc); + if (!p->enc) + return SZ_ERROR_MEM; + } + + limitedInStream.realStream = inStream; + if (inStream) + { + limitedInStream.vt.Read = LimitedSeqInStream_Read; + } + + if (!outBuf) + { + // outStream version works only in one thread. So we use CLzma2Enc::tempBufLzma + if (!me->tempBufLzma) + { + me->tempBufLzma = (Byte *)ISzAlloc_Alloc(me->alloc, LZMA2_CHUNK_SIZE_COMPRESSED_MAX); + if (!me->tempBufLzma) + return SZ_ERROR_MEM; + } + } + + RINOK(Lzma2EncInt_InitStream(p, &me->props)); + + for (;;) + { + SRes res = SZ_OK; + size_t inSizeCur = 0; + + Lzma2EncInt_InitBlock(p); + + LimitedSeqInStream_Init(&limitedInStream); + limitedInStream.limit = me->props.blockSize; + + if (inStream) + { + UInt64 expected = (UInt64)(Int64)-1; + // inStream version works only in one thread. So we use CLzma2Enc::expectedDataSize + if (me->expectedDataSize != (UInt64)(Int64)-1 + && me->expectedDataSize >= unpackTotal) + expected = me->expectedDataSize - unpackTotal; + if (me->props.blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID + && expected > me->props.blockSize) + expected = (size_t)me->props.blockSize; + + LzmaEnc_SetDataSize(p->enc, expected); + + RINOK(LzmaEnc_PrepareForLzma2(p->enc, + &limitedInStream.vt, + LZMA2_KEEP_WINDOW_SIZE, + me->alloc, + me->allocBig)); + } + else + { + inSizeCur = inDataSize - (size_t)unpackTotal; + if (me->props.blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID + && inSizeCur > me->props.blockSize) + inSizeCur = (size_t)me->props.blockSize; + + // LzmaEnc_SetDataSize(p->enc, inSizeCur); + + RINOK(LzmaEnc_MemPrepare(p->enc, + inData + (size_t)unpackTotal, inSizeCur, + LZMA2_KEEP_WINDOW_SIZE, + me->alloc, + me->allocBig)); + } + + for (;;) + { + size_t packSize = LZMA2_CHUNK_SIZE_COMPRESSED_MAX; + if (outBuf) + packSize = outLim - (size_t)packTotal; + + res = Lzma2EncInt_EncodeSubblock(p, + outBuf ? outBuf + (size_t)packTotal : me->tempBufLzma, &packSize, + outBuf ? NULL : outStream); + + if (res != SZ_OK) + break; + + packTotal += packSize; + if (outBuf) + *outBufSize = (size_t)packTotal; + + res = Progress(progress, unpackTotal + p->srcPos, packTotal); + if (res != SZ_OK) + break; + + /* + if (LzmaEnc_GetNumAvailableBytes(p->enc) == 0) + break; + */ + + if (packSize == 0) + break; + } + + LzmaEnc_Finish(p->enc); + + unpackTotal += p->srcPos; + + RINOK(res); + + if (p->srcPos != (inStream ? limitedInStream.processed : inSizeCur)) + return SZ_ERROR_FAIL; + + if (inStream ? limitedInStream.finished : (unpackTotal == inDataSize)) + { + if (finished) + { + if (outBuf) + { + const size_t destPos = *outBufSize; + if (destPos >= outLim) + return SZ_ERROR_OUTPUT_EOF; + outBuf[destPos] = LZMA2_CONTROL_EOF; // 0 + *outBufSize = destPos + 1; + } + else + { + const Byte b = LZMA2_CONTROL_EOF; // 0; + if (ISeqOutStream_Write(outStream, &b, 1) != 1) + return SZ_ERROR_WRITE; + } + } + return SZ_OK; + } + } +} + + + +#ifndef _7ZIP_ST + +static SRes Lzma2Enc_MtCallback_Code(void *pp, unsigned coderIndex, unsigned outBufIndex, + const Byte *src, size_t srcSize, int finished) +{ + CLzma2Enc *me = (CLzma2Enc *)pp; + size_t destSize = me->outBufSize; + SRes res; + CMtProgressThunk progressThunk; + + Byte *dest = me->outBufs[outBufIndex]; + + me->outBufsDataSizes[outBufIndex] = 0; + + if (!dest) + { + dest = (Byte *)ISzAlloc_Alloc(me->alloc, me->outBufSize); + if (!dest) + return SZ_ERROR_MEM; + me->outBufs[outBufIndex] = dest; + } + + MtProgressThunk_CreateVTable(&progressThunk); + progressThunk.mtProgress = &me->mtCoder.mtProgress; + progressThunk.inSize = 0; + progressThunk.outSize = 0; + + res = Lzma2Enc_EncodeMt1(me, + &me->coders[coderIndex], + NULL, dest, &destSize, + NULL, src, srcSize, + finished, + &progressThunk.vt); + + me->outBufsDataSizes[outBufIndex] = destSize; + + return res; +} + + +static SRes Lzma2Enc_MtCallback_Write(void *pp, unsigned outBufIndex) +{ + CLzma2Enc *me = (CLzma2Enc *)pp; + size_t size = me->outBufsDataSizes[outBufIndex]; + const Byte *data = me->outBufs[outBufIndex]; + + if (me->outStream) + return ISeqOutStream_Write(me->outStream, data, size) == size ? SZ_OK : SZ_ERROR_WRITE; + + if (size > me->outBuf_Rem) + return SZ_ERROR_OUTPUT_EOF; + memcpy(me->outBuf, data, size); + me->outBuf_Rem -= size; + me->outBuf += size; + return SZ_OK; +} + +#endif + + + +SRes Lzma2Enc_Encode2(CLzma2EncHandle pp, + ISeqOutStream *outStream, + Byte *outBuf, size_t *outBufSize, + ISeqInStream *inStream, + const Byte *inData, size_t inDataSize, + ICompressProgress *progress) +{ + CLzma2Enc *p = (CLzma2Enc *)pp; + + if (inStream && inData) + return SZ_ERROR_PARAM; + + if (outStream && outBuf) + return SZ_ERROR_PARAM; + + { + unsigned i; + for (i = 0; i < MTCODER__THREADS_MAX; i++) + p->coders[i].propsAreSet = False; + } + + #ifndef _7ZIP_ST + + if (p->props.numBlockThreads_Reduced > 1) + { + IMtCoderCallback2 vt; + + if (!p->mtCoder_WasConstructed) + { + p->mtCoder_WasConstructed = True; + MtCoder_Construct(&p->mtCoder); + } + + vt.Code = Lzma2Enc_MtCallback_Code; + vt.Write = Lzma2Enc_MtCallback_Write; + + p->outStream = outStream; + p->outBuf = NULL; + p->outBuf_Rem = 0; + if (!outStream) + { + p->outBuf = outBuf; + p->outBuf_Rem = *outBufSize; + *outBufSize = 0; + } + + p->mtCoder.allocBig = p->allocBig; + p->mtCoder.progress = progress; + p->mtCoder.inStream = inStream; + p->mtCoder.inData = inData; + p->mtCoder.inDataSize = inDataSize; + p->mtCoder.mtCallback = &vt; + p->mtCoder.mtCallbackObject = p; + + p->mtCoder.blockSize = (size_t)p->props.blockSize; + if (p->mtCoder.blockSize != p->props.blockSize) + return SZ_ERROR_PARAM; /* SZ_ERROR_MEM */ + + { + size_t destBlockSize = p->mtCoder.blockSize + (p->mtCoder.blockSize >> 10) + 16; + if (destBlockSize < p->mtCoder.blockSize) + return SZ_ERROR_PARAM; + if (p->outBufSize != destBlockSize) + Lzma2Enc_FreeOutBufs(p); + p->outBufSize = destBlockSize; + } + + p->mtCoder.numThreadsMax = (unsigned)p->props.numBlockThreads_Max; + p->mtCoder.expectedDataSize = p->expectedDataSize; + + { + SRes res = MtCoder_Code(&p->mtCoder); + if (!outStream) + *outBufSize = (size_t)(p->outBuf - outBuf); + return res; + } + } + + #endif + + + return Lzma2Enc_EncodeMt1(p, + &p->coders[0], + outStream, outBuf, outBufSize, + inStream, inData, inDataSize, + True, /* finished */ + progress); +} diff --git a/deps/LZMA-SDK/C/Lzma2Enc.h b/deps/LZMA-SDK/C/Lzma2Enc.h index 65f2dd145..6a6110ff7 100644 --- a/deps/LZMA-SDK/C/Lzma2Enc.h +++ b/deps/LZMA-SDK/C/Lzma2Enc.h @@ -1,55 +1,55 @@ -/* Lzma2Enc.h -- LZMA2 Encoder -2017-07-27 : Igor Pavlov : Public domain */ - -#ifndef __LZMA2_ENC_H -#define __LZMA2_ENC_H - -#include "LzmaEnc.h" - -EXTERN_C_BEGIN - -#define LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO 0 -#define LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID ((UInt64)(Int64)-1) - -typedef struct -{ - CLzmaEncProps lzmaProps; - UInt64 blockSize; - int numBlockThreads_Reduced; - int numBlockThreads_Max; - int numTotalThreads; -} CLzma2EncProps; - -void Lzma2EncProps_Init(CLzma2EncProps *p); -void Lzma2EncProps_Normalize(CLzma2EncProps *p); - -/* ---------- CLzmaEnc2Handle Interface ---------- */ - -/* Lzma2Enc_* functions can return the following exit codes: -SRes: - SZ_OK - OK - SZ_ERROR_MEM - Memory allocation error - SZ_ERROR_PARAM - Incorrect paramater in props - SZ_ERROR_WRITE - ISeqOutStream write callback error - SZ_ERROR_OUTPUT_EOF - output buffer overflow - version with (Byte *) output - SZ_ERROR_PROGRESS - some break from progress callback - SZ_ERROR_THREAD - error in multithreading functions (only for Mt version) -*/ - -typedef void * CLzma2EncHandle; - -CLzma2EncHandle Lzma2Enc_Create(ISzAllocPtr alloc, ISzAllocPtr allocBig); -void Lzma2Enc_Destroy(CLzma2EncHandle p); -SRes Lzma2Enc_SetProps(CLzma2EncHandle p, const CLzma2EncProps *props); -void Lzma2Enc_SetDataSize(CLzma2EncHandle p, UInt64 expectedDataSiize); -Byte Lzma2Enc_WriteProperties(CLzma2EncHandle p); -SRes Lzma2Enc_Encode2(CLzma2EncHandle p, - ISeqOutStream *outStream, - Byte *outBuf, size_t *outBufSize, - ISeqInStream *inStream, - const Byte *inData, size_t inDataSize, - ICompressProgress *progress); - -EXTERN_C_END - -#endif +/* Lzma2Enc.h -- LZMA2 Encoder +2017-07-27 : Igor Pavlov : Public domain */ + +#ifndef __LZMA2_ENC_H +#define __LZMA2_ENC_H + +#include "LzmaEnc.h" + +EXTERN_C_BEGIN + +#define LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO 0 +#define LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID ((UInt64)(Int64)-1) + +typedef struct +{ + CLzmaEncProps lzmaProps; + UInt64 blockSize; + int numBlockThreads_Reduced; + int numBlockThreads_Max; + int numTotalThreads; +} CLzma2EncProps; + +void Lzma2EncProps_Init(CLzma2EncProps *p); +void Lzma2EncProps_Normalize(CLzma2EncProps *p); + +/* ---------- CLzmaEnc2Handle Interface ---------- */ + +/* Lzma2Enc_* functions can return the following exit codes: +SRes: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater in props + SZ_ERROR_WRITE - ISeqOutStream write callback error + SZ_ERROR_OUTPUT_EOF - output buffer overflow - version with (Byte *) output + SZ_ERROR_PROGRESS - some break from progress callback + SZ_ERROR_THREAD - error in multithreading functions (only for Mt version) +*/ + +typedef void * CLzma2EncHandle; + +CLzma2EncHandle Lzma2Enc_Create(ISzAllocPtr alloc, ISzAllocPtr allocBig); +void Lzma2Enc_Destroy(CLzma2EncHandle p); +SRes Lzma2Enc_SetProps(CLzma2EncHandle p, const CLzma2EncProps *props); +void Lzma2Enc_SetDataSize(CLzma2EncHandle p, UInt64 expectedDataSiize); +Byte Lzma2Enc_WriteProperties(CLzma2EncHandle p); +SRes Lzma2Enc_Encode2(CLzma2EncHandle p, + ISeqOutStream *outStream, + Byte *outBuf, size_t *outBufSize, + ISeqInStream *inStream, + const Byte *inData, size_t inDataSize, + ICompressProgress *progress); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/Lzma86.h b/deps/LZMA-SDK/C/Lzma86.h index 83057e598..bebed5cb7 100644 --- a/deps/LZMA-SDK/C/Lzma86.h +++ b/deps/LZMA-SDK/C/Lzma86.h @@ -1,111 +1,111 @@ -/* Lzma86.h -- LZMA + x86 (BCJ) Filter -2013-01-18 : Igor Pavlov : Public domain */ - -#ifndef __LZMA86_H -#define __LZMA86_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -#define LZMA86_SIZE_OFFSET (1 + 5) -#define LZMA86_HEADER_SIZE (LZMA86_SIZE_OFFSET + 8) - -/* -It's an example for LZMA + x86 Filter use. -You can use .lzma86 extension, if you write that stream to file. -.lzma86 header adds one additional byte to standard .lzma header. -.lzma86 header (14 bytes): - Offset Size Description - 0 1 = 0 - no filter, pure LZMA - = 1 - x86 filter + LZMA - 1 1 lc, lp and pb in encoded form - 2 4 dictSize (little endian) - 6 8 uncompressed size (little endian) - - -Lzma86_Encode -------------- -level - compression level: 0 <= level <= 9, the default value for "level" is 5. - -dictSize - The dictionary size in bytes. The maximum value is - 128 MB = (1 << 27) bytes for 32-bit version - 1 GB = (1 << 30) bytes for 64-bit version - The default value is 16 MB = (1 << 24) bytes, for level = 5. - It's recommended to use the dictionary that is larger than 4 KB and - that can be calculated as (1 << N) or (3 << N) sizes. - For better compression ratio dictSize must be >= inSize. - -filterMode: - SZ_FILTER_NO - no Filter - SZ_FILTER_YES - x86 Filter - SZ_FILTER_AUTO - it tries both alternatives to select best. - Encoder will use 2 or 3 passes: - 2 passes when FILTER_NO provides better compression. - 3 passes when FILTER_YES provides better compression. - -Lzma86Encode allocates Data with MyAlloc functions. -RAM Requirements for compressing: - RamSize = dictionarySize * 11.5 + 6MB + FilterBlockSize - filterMode FilterBlockSize - SZ_FILTER_NO 0 - SZ_FILTER_YES inSize - SZ_FILTER_AUTO inSize - - -Return code: - SZ_OK - OK - SZ_ERROR_MEM - Memory allocation error - SZ_ERROR_PARAM - Incorrect paramater - SZ_ERROR_OUTPUT_EOF - output buffer overflow - SZ_ERROR_THREAD - errors in multithreading functions (only for Mt version) -*/ - -enum ESzFilterMode -{ - SZ_FILTER_NO, - SZ_FILTER_YES, - SZ_FILTER_AUTO -}; - -SRes Lzma86_Encode(Byte *dest, size_t *destLen, const Byte *src, size_t srcLen, - int level, UInt32 dictSize, int filterMode); - - -/* -Lzma86_GetUnpackSize: - In: - src - input data - srcLen - input data size - Out: - unpackSize - size of uncompressed stream - Return code: - SZ_OK - OK - SZ_ERROR_INPUT_EOF - Error in headers -*/ - -SRes Lzma86_GetUnpackSize(const Byte *src, SizeT srcLen, UInt64 *unpackSize); - -/* -Lzma86_Decode: - In: - dest - output data - destLen - output data size - src - input data - srcLen - input data size - Out: - destLen - processed output size - srcLen - processed input size - Return code: - SZ_OK - OK - SZ_ERROR_DATA - Data error - SZ_ERROR_MEM - Memory allocation error - SZ_ERROR_UNSUPPORTED - unsupported file - SZ_ERROR_INPUT_EOF - it needs more bytes in input buffer -*/ - -SRes Lzma86_Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen); - -EXTERN_C_END - -#endif +/* Lzma86.h -- LZMA + x86 (BCJ) Filter +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __LZMA86_H +#define __LZMA86_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define LZMA86_SIZE_OFFSET (1 + 5) +#define LZMA86_HEADER_SIZE (LZMA86_SIZE_OFFSET + 8) + +/* +It's an example for LZMA + x86 Filter use. +You can use .lzma86 extension, if you write that stream to file. +.lzma86 header adds one additional byte to standard .lzma header. +.lzma86 header (14 bytes): + Offset Size Description + 0 1 = 0 - no filter, pure LZMA + = 1 - x86 filter + LZMA + 1 1 lc, lp and pb in encoded form + 2 4 dictSize (little endian) + 6 8 uncompressed size (little endian) + + +Lzma86_Encode +------------- +level - compression level: 0 <= level <= 9, the default value for "level" is 5. + +dictSize - The dictionary size in bytes. The maximum value is + 128 MB = (1 << 27) bytes for 32-bit version + 1 GB = (1 << 30) bytes for 64-bit version + The default value is 16 MB = (1 << 24) bytes, for level = 5. + It's recommended to use the dictionary that is larger than 4 KB and + that can be calculated as (1 << N) or (3 << N) sizes. + For better compression ratio dictSize must be >= inSize. + +filterMode: + SZ_FILTER_NO - no Filter + SZ_FILTER_YES - x86 Filter + SZ_FILTER_AUTO - it tries both alternatives to select best. + Encoder will use 2 or 3 passes: + 2 passes when FILTER_NO provides better compression. + 3 passes when FILTER_YES provides better compression. + +Lzma86Encode allocates Data with MyAlloc functions. +RAM Requirements for compressing: + RamSize = dictionarySize * 11.5 + 6MB + FilterBlockSize + filterMode FilterBlockSize + SZ_FILTER_NO 0 + SZ_FILTER_YES inSize + SZ_FILTER_AUTO inSize + + +Return code: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater + SZ_ERROR_OUTPUT_EOF - output buffer overflow + SZ_ERROR_THREAD - errors in multithreading functions (only for Mt version) +*/ + +enum ESzFilterMode +{ + SZ_FILTER_NO, + SZ_FILTER_YES, + SZ_FILTER_AUTO +}; + +SRes Lzma86_Encode(Byte *dest, size_t *destLen, const Byte *src, size_t srcLen, + int level, UInt32 dictSize, int filterMode); + + +/* +Lzma86_GetUnpackSize: + In: + src - input data + srcLen - input data size + Out: + unpackSize - size of uncompressed stream + Return code: + SZ_OK - OK + SZ_ERROR_INPUT_EOF - Error in headers +*/ + +SRes Lzma86_GetUnpackSize(const Byte *src, SizeT srcLen, UInt64 *unpackSize); + +/* +Lzma86_Decode: + In: + dest - output data + destLen - output data size + src - input data + srcLen - input data size + Out: + destLen - processed output size + srcLen - processed input size + Return code: + SZ_OK - OK + SZ_ERROR_DATA - Data error + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_UNSUPPORTED - unsupported file + SZ_ERROR_INPUT_EOF - it needs more bytes in input buffer +*/ + +SRes Lzma86_Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/Lzma86Dec.c b/deps/LZMA-SDK/C/Lzma86Dec.c index 20ac5e7a9..21031745c 100644 --- a/deps/LZMA-SDK/C/Lzma86Dec.c +++ b/deps/LZMA-SDK/C/Lzma86Dec.c @@ -1,54 +1,54 @@ -/* Lzma86Dec.c -- LZMA + x86 (BCJ) Filter Decoder -2016-05-16 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "Lzma86.h" - -#include "Alloc.h" -#include "Bra.h" -#include "LzmaDec.h" - -SRes Lzma86_GetUnpackSize(const Byte *src, SizeT srcLen, UInt64 *unpackSize) -{ - unsigned i; - if (srcLen < LZMA86_HEADER_SIZE) - return SZ_ERROR_INPUT_EOF; - *unpackSize = 0; - for (i = 0; i < sizeof(UInt64); i++) - *unpackSize += ((UInt64)src[LZMA86_SIZE_OFFSET + i]) << (8 * i); - return SZ_OK; -} - -SRes Lzma86_Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen) -{ - SRes res; - int useFilter; - SizeT inSizePure; - ELzmaStatus status; - - if (*srcLen < LZMA86_HEADER_SIZE) - return SZ_ERROR_INPUT_EOF; - - useFilter = src[0]; - - if (useFilter > 1) - { - *destLen = 0; - return SZ_ERROR_UNSUPPORTED; - } - - inSizePure = *srcLen - LZMA86_HEADER_SIZE; - res = LzmaDecode(dest, destLen, src + LZMA86_HEADER_SIZE, &inSizePure, - src + 1, LZMA_PROPS_SIZE, LZMA_FINISH_ANY, &status, &g_Alloc); - *srcLen = inSizePure + LZMA86_HEADER_SIZE; - if (res != SZ_OK) - return res; - if (useFilter == 1) - { - UInt32 x86State; - x86_Convert_Init(x86State); - x86_Convert(dest, *destLen, 0, &x86State, 0); - } - return SZ_OK; -} +/* Lzma86Dec.c -- LZMA + x86 (BCJ) Filter Decoder +2016-05-16 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "Lzma86.h" + +#include "Alloc.h" +#include "Bra.h" +#include "LzmaDec.h" + +SRes Lzma86_GetUnpackSize(const Byte *src, SizeT srcLen, UInt64 *unpackSize) +{ + unsigned i; + if (srcLen < LZMA86_HEADER_SIZE) + return SZ_ERROR_INPUT_EOF; + *unpackSize = 0; + for (i = 0; i < sizeof(UInt64); i++) + *unpackSize += ((UInt64)src[LZMA86_SIZE_OFFSET + i]) << (8 * i); + return SZ_OK; +} + +SRes Lzma86_Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen) +{ + SRes res; + int useFilter; + SizeT inSizePure; + ELzmaStatus status; + + if (*srcLen < LZMA86_HEADER_SIZE) + return SZ_ERROR_INPUT_EOF; + + useFilter = src[0]; + + if (useFilter > 1) + { + *destLen = 0; + return SZ_ERROR_UNSUPPORTED; + } + + inSizePure = *srcLen - LZMA86_HEADER_SIZE; + res = LzmaDecode(dest, destLen, src + LZMA86_HEADER_SIZE, &inSizePure, + src + 1, LZMA_PROPS_SIZE, LZMA_FINISH_ANY, &status, &g_Alloc); + *srcLen = inSizePure + LZMA86_HEADER_SIZE; + if (res != SZ_OK) + return res; + if (useFilter == 1) + { + UInt32 x86State; + x86_Convert_Init(x86State); + x86_Convert(dest, *destLen, 0, &x86State, 0); + } + return SZ_OK; +} diff --git a/deps/LZMA-SDK/C/Lzma86Enc.c b/deps/LZMA-SDK/C/Lzma86Enc.c index 99397bc5e..14fcd65c6 100644 --- a/deps/LZMA-SDK/C/Lzma86Enc.c +++ b/deps/LZMA-SDK/C/Lzma86Enc.c @@ -1,104 +1,104 @@ -/* Lzma86Enc.c -- LZMA + x86 (BCJ) Filter Encoder -2018-07-04 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include - -#include "Lzma86.h" - -#include "Alloc.h" -#include "Bra.h" -#include "LzmaEnc.h" - -int Lzma86_Encode(Byte *dest, size_t *destLen, const Byte *src, size_t srcLen, - int level, UInt32 dictSize, int filterMode) -{ - size_t outSize2 = *destLen; - Byte *filteredStream; - BoolInt useFilter; - int mainResult = SZ_ERROR_OUTPUT_EOF; - CLzmaEncProps props; - LzmaEncProps_Init(&props); - props.level = level; - props.dictSize = dictSize; - - *destLen = 0; - if (outSize2 < LZMA86_HEADER_SIZE) - return SZ_ERROR_OUTPUT_EOF; - - { - int i; - UInt64 t = srcLen; - for (i = 0; i < 8; i++, t >>= 8) - dest[LZMA86_SIZE_OFFSET + i] = (Byte)t; - } - - filteredStream = 0; - useFilter = (filterMode != SZ_FILTER_NO); - if (useFilter) - { - if (srcLen != 0) - { - filteredStream = (Byte *)MyAlloc(srcLen); - if (filteredStream == 0) - return SZ_ERROR_MEM; - memcpy(filteredStream, src, srcLen); - } - { - UInt32 x86State; - x86_Convert_Init(x86State); - x86_Convert(filteredStream, srcLen, 0, &x86State, 1); - } - } - - { - size_t minSize = 0; - BoolInt bestIsFiltered = False; - - /* passes for SZ_FILTER_AUTO: - 0 - BCJ + LZMA - 1 - LZMA - 2 - BCJ + LZMA agaian, if pass 0 (BCJ + LZMA) is better. - */ - int numPasses = (filterMode == SZ_FILTER_AUTO) ? 3 : 1; - - int i; - for (i = 0; i < numPasses; i++) - { - size_t outSizeProcessed = outSize2 - LZMA86_HEADER_SIZE; - size_t outPropsSize = 5; - SRes curRes; - BoolInt curModeIsFiltered = (numPasses > 1 && i == numPasses - 1); - if (curModeIsFiltered && !bestIsFiltered) - break; - if (useFilter && i == 0) - curModeIsFiltered = True; - - curRes = LzmaEncode(dest + LZMA86_HEADER_SIZE, &outSizeProcessed, - curModeIsFiltered ? filteredStream : src, srcLen, - &props, dest + 1, &outPropsSize, 0, - NULL, &g_Alloc, &g_Alloc); - - if (curRes != SZ_ERROR_OUTPUT_EOF) - { - if (curRes != SZ_OK) - { - mainResult = curRes; - break; - } - if (outSizeProcessed <= minSize || mainResult != SZ_OK) - { - minSize = outSizeProcessed; - bestIsFiltered = curModeIsFiltered; - mainResult = SZ_OK; - } - } - } - dest[0] = (Byte)(bestIsFiltered ? 1 : 0); - *destLen = LZMA86_HEADER_SIZE + minSize; - } - if (useFilter) - MyFree(filteredStream); - return mainResult; -} +/* Lzma86Enc.c -- LZMA + x86 (BCJ) Filter Encoder +2018-07-04 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#include "Lzma86.h" + +#include "Alloc.h" +#include "Bra.h" +#include "LzmaEnc.h" + +int Lzma86_Encode(Byte *dest, size_t *destLen, const Byte *src, size_t srcLen, + int level, UInt32 dictSize, int filterMode) +{ + size_t outSize2 = *destLen; + Byte *filteredStream; + BoolInt useFilter; + int mainResult = SZ_ERROR_OUTPUT_EOF; + CLzmaEncProps props; + LzmaEncProps_Init(&props); + props.level = level; + props.dictSize = dictSize; + + *destLen = 0; + if (outSize2 < LZMA86_HEADER_SIZE) + return SZ_ERROR_OUTPUT_EOF; + + { + int i; + UInt64 t = srcLen; + for (i = 0; i < 8; i++, t >>= 8) + dest[LZMA86_SIZE_OFFSET + i] = (Byte)t; + } + + filteredStream = 0; + useFilter = (filterMode != SZ_FILTER_NO); + if (useFilter) + { + if (srcLen != 0) + { + filteredStream = (Byte *)MyAlloc(srcLen); + if (filteredStream == 0) + return SZ_ERROR_MEM; + memcpy(filteredStream, src, srcLen); + } + { + UInt32 x86State; + x86_Convert_Init(x86State); + x86_Convert(filteredStream, srcLen, 0, &x86State, 1); + } + } + + { + size_t minSize = 0; + BoolInt bestIsFiltered = False; + + /* passes for SZ_FILTER_AUTO: + 0 - BCJ + LZMA + 1 - LZMA + 2 - BCJ + LZMA agaian, if pass 0 (BCJ + LZMA) is better. + */ + int numPasses = (filterMode == SZ_FILTER_AUTO) ? 3 : 1; + + int i; + for (i = 0; i < numPasses; i++) + { + size_t outSizeProcessed = outSize2 - LZMA86_HEADER_SIZE; + size_t outPropsSize = 5; + SRes curRes; + BoolInt curModeIsFiltered = (numPasses > 1 && i == numPasses - 1); + if (curModeIsFiltered && !bestIsFiltered) + break; + if (useFilter && i == 0) + curModeIsFiltered = True; + + curRes = LzmaEncode(dest + LZMA86_HEADER_SIZE, &outSizeProcessed, + curModeIsFiltered ? filteredStream : src, srcLen, + &props, dest + 1, &outPropsSize, 0, + NULL, &g_Alloc, &g_Alloc); + + if (curRes != SZ_ERROR_OUTPUT_EOF) + { + if (curRes != SZ_OK) + { + mainResult = curRes; + break; + } + if (outSizeProcessed <= minSize || mainResult != SZ_OK) + { + minSize = outSizeProcessed; + bestIsFiltered = curModeIsFiltered; + mainResult = SZ_OK; + } + } + } + dest[0] = (Byte)(bestIsFiltered ? 1 : 0); + *destLen = LZMA86_HEADER_SIZE + minSize; + } + if (useFilter) + MyFree(filteredStream); + return mainResult; +} diff --git a/deps/LZMA-SDK/C/LzmaDec.c b/deps/LZMA-SDK/C/LzmaDec.c index 80b70a9ee..d6742e5af 100644 --- a/deps/LZMA-SDK/C/LzmaDec.c +++ b/deps/LZMA-SDK/C/LzmaDec.c @@ -1,1363 +1,1363 @@ -/* LzmaDec.c -- LZMA Decoder -2021-04-01 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include - -/* #include "CpuArch.h" */ -#include "LzmaDec.h" - -#define kNumTopBits 24 -#define kTopValue ((UInt32)1 << kNumTopBits) - -#define kNumBitModelTotalBits 11 -#define kBitModelTotal (1 << kNumBitModelTotalBits) - -#define RC_INIT_SIZE 5 - -#ifndef _LZMA_DEC_OPT - -#define kNumMoveBits 5 -#define NORMALIZE if (range < kTopValue) { range <<= 8; code = (code << 8) | (*buf++); } - -#define IF_BIT_0(p) ttt = *(p); NORMALIZE; bound = (range >> kNumBitModelTotalBits) * (UInt32)ttt; if (code < bound) -#define UPDATE_0(p) range = bound; *(p) = (CLzmaProb)(ttt + ((kBitModelTotal - ttt) >> kNumMoveBits)); -#define UPDATE_1(p) range -= bound; code -= bound; *(p) = (CLzmaProb)(ttt - (ttt >> kNumMoveBits)); -#define GET_BIT2(p, i, A0, A1) IF_BIT_0(p) \ - { UPDATE_0(p); i = (i + i); A0; } else \ - { UPDATE_1(p); i = (i + i) + 1; A1; } - -#define TREE_GET_BIT(probs, i) { GET_BIT2(probs + i, i, ;, ;); } - -#define REV_BIT(p, i, A0, A1) IF_BIT_0(p + i) \ - { UPDATE_0(p + i); A0; } else \ - { UPDATE_1(p + i); A1; } -#define REV_BIT_VAR( p, i, m) REV_BIT(p, i, i += m; m += m, m += m; i += m; ) -#define REV_BIT_CONST(p, i, m) REV_BIT(p, i, i += m; , i += m * 2; ) -#define REV_BIT_LAST( p, i, m) REV_BIT(p, i, i -= m , ; ) - -#define TREE_DECODE(probs, limit, i) \ - { i = 1; do { TREE_GET_BIT(probs, i); } while (i < limit); i -= limit; } - -/* #define _LZMA_SIZE_OPT */ - -#ifdef _LZMA_SIZE_OPT -#define TREE_6_DECODE(probs, i) TREE_DECODE(probs, (1 << 6), i) -#else -#define TREE_6_DECODE(probs, i) \ - { i = 1; \ - TREE_GET_BIT(probs, i); \ - TREE_GET_BIT(probs, i); \ - TREE_GET_BIT(probs, i); \ - TREE_GET_BIT(probs, i); \ - TREE_GET_BIT(probs, i); \ - TREE_GET_BIT(probs, i); \ - i -= 0x40; } -#endif - -#define NORMAL_LITER_DEC TREE_GET_BIT(prob, symbol) -#define MATCHED_LITER_DEC \ - matchByte += matchByte; \ - bit = offs; \ - offs &= matchByte; \ - probLit = prob + (offs + bit + symbol); \ - GET_BIT2(probLit, symbol, offs ^= bit; , ;) - -#endif // _LZMA_DEC_OPT - - -#define NORMALIZE_CHECK if (range < kTopValue) { if (buf >= bufLimit) return DUMMY_INPUT_EOF; range <<= 8; code = (code << 8) | (*buf++); } - -#define IF_BIT_0_CHECK(p) ttt = *(p); NORMALIZE_CHECK; bound = (range >> kNumBitModelTotalBits) * (UInt32)ttt; if (code < bound) -#define UPDATE_0_CHECK range = bound; -#define UPDATE_1_CHECK range -= bound; code -= bound; -#define GET_BIT2_CHECK(p, i, A0, A1) IF_BIT_0_CHECK(p) \ - { UPDATE_0_CHECK; i = (i + i); A0; } else \ - { UPDATE_1_CHECK; i = (i + i) + 1; A1; } -#define GET_BIT_CHECK(p, i) GET_BIT2_CHECK(p, i, ; , ;) -#define TREE_DECODE_CHECK(probs, limit, i) \ - { i = 1; do { GET_BIT_CHECK(probs + i, i) } while (i < limit); i -= limit; } - - -#define REV_BIT_CHECK(p, i, m) IF_BIT_0_CHECK(p + i) \ - { UPDATE_0_CHECK; i += m; m += m; } else \ - { UPDATE_1_CHECK; m += m; i += m; } - - -#define kNumPosBitsMax 4 -#define kNumPosStatesMax (1 << kNumPosBitsMax) - -#define kLenNumLowBits 3 -#define kLenNumLowSymbols (1 << kLenNumLowBits) -#define kLenNumHighBits 8 -#define kLenNumHighSymbols (1 << kLenNumHighBits) - -#define LenLow 0 -#define LenHigh (LenLow + 2 * (kNumPosStatesMax << kLenNumLowBits)) -#define kNumLenProbs (LenHigh + kLenNumHighSymbols) - -#define LenChoice LenLow -#define LenChoice2 (LenLow + (1 << kLenNumLowBits)) - -#define kNumStates 12 -#define kNumStates2 16 -#define kNumLitStates 7 - -#define kStartPosModelIndex 4 -#define kEndPosModelIndex 14 -#define kNumFullDistances (1 << (kEndPosModelIndex >> 1)) - -#define kNumPosSlotBits 6 -#define kNumLenToPosStates 4 - -#define kNumAlignBits 4 -#define kAlignTableSize (1 << kNumAlignBits) - -#define kMatchMinLen 2 -#define kMatchSpecLenStart (kMatchMinLen + kLenNumLowSymbols * 2 + kLenNumHighSymbols) - -#define kMatchSpecLen_Error_Data (1 << 9) -#define kMatchSpecLen_Error_Fail (kMatchSpecLen_Error_Data - 1) - -/* External ASM code needs same CLzmaProb array layout. So don't change it. */ - -/* (probs_1664) is faster and better for code size at some platforms */ -/* -#ifdef MY_CPU_X86_OR_AMD64 -*/ -#define kStartOffset 1664 -#define GET_PROBS p->probs_1664 -/* -#define GET_PROBS p->probs + kStartOffset -#else -#define kStartOffset 0 -#define GET_PROBS p->probs -#endif -*/ - -#define SpecPos (-kStartOffset) -#define IsRep0Long (SpecPos + kNumFullDistances) -#define RepLenCoder (IsRep0Long + (kNumStates2 << kNumPosBitsMax)) -#define LenCoder (RepLenCoder + kNumLenProbs) -#define IsMatch (LenCoder + kNumLenProbs) -#define Align (IsMatch + (kNumStates2 << kNumPosBitsMax)) -#define IsRep (Align + kAlignTableSize) -#define IsRepG0 (IsRep + kNumStates) -#define IsRepG1 (IsRepG0 + kNumStates) -#define IsRepG2 (IsRepG1 + kNumStates) -#define PosSlot (IsRepG2 + kNumStates) -#define Literal (PosSlot + (kNumLenToPosStates << kNumPosSlotBits)) -#define NUM_BASE_PROBS (Literal + kStartOffset) - -#if Align != 0 && kStartOffset != 0 - #error Stop_Compiling_Bad_LZMA_kAlign -#endif - -#if NUM_BASE_PROBS != 1984 - #error Stop_Compiling_Bad_LZMA_PROBS -#endif - - -#define LZMA_LIT_SIZE 0x300 - -#define LzmaProps_GetNumProbs(p) (NUM_BASE_PROBS + ((UInt32)LZMA_LIT_SIZE << ((p)->lc + (p)->lp))) - - -#define CALC_POS_STATE(processedPos, pbMask) (((processedPos) & (pbMask)) << 4) -#define COMBINED_PS_STATE (posState + state) -#define GET_LEN_STATE (posState) - -#define LZMA_DIC_MIN (1 << 12) - -/* -p->remainLen : shows status of LZMA decoder: - < kMatchSpecLenStart : the number of bytes to be copied with (p->rep0) offset - = kMatchSpecLenStart : the LZMA stream was finished with end mark - = kMatchSpecLenStart + 1 : need init range coder - = kMatchSpecLenStart + 2 : need init range coder and state - = kMatchSpecLen_Error_Fail : Internal Code Failure - = kMatchSpecLen_Error_Data + [0 ... 273] : LZMA Data Error -*/ - -/* ---------- LZMA_DECODE_REAL ---------- */ -/* -LzmaDec_DecodeReal_3() can be implemented in external ASM file. -3 - is the code compatibility version of that function for check at link time. -*/ - -#define LZMA_DECODE_REAL LzmaDec_DecodeReal_3 - -/* -LZMA_DECODE_REAL() -In: - RangeCoder is normalized - if (p->dicPos == limit) - { - LzmaDec_TryDummy() was called before to exclude LITERAL and MATCH-REP cases. - So first symbol can be only MATCH-NON-REP. And if that MATCH-NON-REP symbol - is not END_OF_PAYALOAD_MARKER, then the function doesn't write any byte to dictionary, - the function returns SZ_OK, and the caller can use (p->remainLen) and (p->reps[0]) later. - } - -Processing: - The first LZMA symbol will be decoded in any case. - All main checks for limits are at the end of main loop, - It decodes additional LZMA-symbols while (p->buf < bufLimit && dicPos < limit), - RangeCoder is still without last normalization when (p->buf < bufLimit) is being checked. - But if (p->buf < bufLimit), the caller provided at least (LZMA_REQUIRED_INPUT_MAX + 1) bytes for - next iteration before limit (bufLimit + LZMA_REQUIRED_INPUT_MAX), - that is enough for worst case LZMA symbol with one additional RangeCoder normalization for one bit. - So that function never reads bufLimit [LZMA_REQUIRED_INPUT_MAX] byte. - -Out: - RangeCoder is normalized - Result: - SZ_OK - OK - p->remainLen: - < kMatchSpecLenStart : the number of bytes to be copied with (p->reps[0]) offset - = kMatchSpecLenStart : the LZMA stream was finished with end mark - - SZ_ERROR_DATA - error, when the MATCH-Symbol refers out of dictionary - p->remainLen : undefined - p->reps[*] : undefined -*/ - - -#ifdef _LZMA_DEC_OPT - -int MY_FAST_CALL LZMA_DECODE_REAL(CLzmaDec *p, SizeT limit, const Byte *bufLimit); - -#else - -static -int MY_FAST_CALL LZMA_DECODE_REAL(CLzmaDec *p, SizeT limit, const Byte *bufLimit) -{ - CLzmaProb *probs = GET_PROBS; - unsigned state = (unsigned)p->state; - UInt32 rep0 = p->reps[0], rep1 = p->reps[1], rep2 = p->reps[2], rep3 = p->reps[3]; - unsigned pbMask = ((unsigned)1 << (p->prop.pb)) - 1; - unsigned lc = p->prop.lc; - unsigned lpMask = ((unsigned)0x100 << p->prop.lp) - ((unsigned)0x100 >> lc); - - Byte *dic = p->dic; - SizeT dicBufSize = p->dicBufSize; - SizeT dicPos = p->dicPos; - - UInt32 processedPos = p->processedPos; - UInt32 checkDicSize = p->checkDicSize; - unsigned len = 0; - - const Byte *buf = p->buf; - UInt32 range = p->range; - UInt32 code = p->code; - - do - { - CLzmaProb *prob; - UInt32 bound; - unsigned ttt; - unsigned posState = CALC_POS_STATE(processedPos, pbMask); - - prob = probs + IsMatch + COMBINED_PS_STATE; - IF_BIT_0(prob) - { - unsigned symbol; - UPDATE_0(prob); - prob = probs + Literal; - if (processedPos != 0 || checkDicSize != 0) - prob += (UInt32)3 * ((((processedPos << 8) + dic[(dicPos == 0 ? dicBufSize : dicPos) - 1]) & lpMask) << lc); - processedPos++; - - if (state < kNumLitStates) - { - state -= (state < 4) ? state : 3; - symbol = 1; - #ifdef _LZMA_SIZE_OPT - do { NORMAL_LITER_DEC } while (symbol < 0x100); - #else - NORMAL_LITER_DEC - NORMAL_LITER_DEC - NORMAL_LITER_DEC - NORMAL_LITER_DEC - NORMAL_LITER_DEC - NORMAL_LITER_DEC - NORMAL_LITER_DEC - NORMAL_LITER_DEC - #endif - } - else - { - unsigned matchByte = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)]; - unsigned offs = 0x100; - state -= (state < 10) ? 3 : 6; - symbol = 1; - #ifdef _LZMA_SIZE_OPT - do - { - unsigned bit; - CLzmaProb *probLit; - MATCHED_LITER_DEC - } - while (symbol < 0x100); - #else - { - unsigned bit; - CLzmaProb *probLit; - MATCHED_LITER_DEC - MATCHED_LITER_DEC - MATCHED_LITER_DEC - MATCHED_LITER_DEC - MATCHED_LITER_DEC - MATCHED_LITER_DEC - MATCHED_LITER_DEC - MATCHED_LITER_DEC - } - #endif - } - - dic[dicPos++] = (Byte)symbol; - continue; - } - - { - UPDATE_1(prob); - prob = probs + IsRep + state; - IF_BIT_0(prob) - { - UPDATE_0(prob); - state += kNumStates; - prob = probs + LenCoder; - } - else - { - UPDATE_1(prob); - prob = probs + IsRepG0 + state; - IF_BIT_0(prob) - { - UPDATE_0(prob); - prob = probs + IsRep0Long + COMBINED_PS_STATE; - IF_BIT_0(prob) - { - UPDATE_0(prob); - - // that case was checked before with kBadRepCode - // if (checkDicSize == 0 && processedPos == 0) { len = kMatchSpecLen_Error_Data + 1; break; } - // The caller doesn't allow (dicPos == limit) case here - // so we don't need the following check: - // if (dicPos == limit) { state = state < kNumLitStates ? 9 : 11; len = 1; break; } - - dic[dicPos] = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)]; - dicPos++; - processedPos++; - state = state < kNumLitStates ? 9 : 11; - continue; - } - UPDATE_1(prob); - } - else - { - UInt32 distance; - UPDATE_1(prob); - prob = probs + IsRepG1 + state; - IF_BIT_0(prob) - { - UPDATE_0(prob); - distance = rep1; - } - else - { - UPDATE_1(prob); - prob = probs + IsRepG2 + state; - IF_BIT_0(prob) - { - UPDATE_0(prob); - distance = rep2; - } - else - { - UPDATE_1(prob); - distance = rep3; - rep3 = rep2; - } - rep2 = rep1; - } - rep1 = rep0; - rep0 = distance; - } - state = state < kNumLitStates ? 8 : 11; - prob = probs + RepLenCoder; - } - - #ifdef _LZMA_SIZE_OPT - { - unsigned lim, offset; - CLzmaProb *probLen = prob + LenChoice; - IF_BIT_0(probLen) - { - UPDATE_0(probLen); - probLen = prob + LenLow + GET_LEN_STATE; - offset = 0; - lim = (1 << kLenNumLowBits); - } - else - { - UPDATE_1(probLen); - probLen = prob + LenChoice2; - IF_BIT_0(probLen) - { - UPDATE_0(probLen); - probLen = prob + LenLow + GET_LEN_STATE + (1 << kLenNumLowBits); - offset = kLenNumLowSymbols; - lim = (1 << kLenNumLowBits); - } - else - { - UPDATE_1(probLen); - probLen = prob + LenHigh; - offset = kLenNumLowSymbols * 2; - lim = (1 << kLenNumHighBits); - } - } - TREE_DECODE(probLen, lim, len); - len += offset; - } - #else - { - CLzmaProb *probLen = prob + LenChoice; - IF_BIT_0(probLen) - { - UPDATE_0(probLen); - probLen = prob + LenLow + GET_LEN_STATE; - len = 1; - TREE_GET_BIT(probLen, len); - TREE_GET_BIT(probLen, len); - TREE_GET_BIT(probLen, len); - len -= 8; - } - else - { - UPDATE_1(probLen); - probLen = prob + LenChoice2; - IF_BIT_0(probLen) - { - UPDATE_0(probLen); - probLen = prob + LenLow + GET_LEN_STATE + (1 << kLenNumLowBits); - len = 1; - TREE_GET_BIT(probLen, len); - TREE_GET_BIT(probLen, len); - TREE_GET_BIT(probLen, len); - } - else - { - UPDATE_1(probLen); - probLen = prob + LenHigh; - TREE_DECODE(probLen, (1 << kLenNumHighBits), len); - len += kLenNumLowSymbols * 2; - } - } - } - #endif - - if (state >= kNumStates) - { - UInt32 distance; - prob = probs + PosSlot + - ((len < kNumLenToPosStates ? len : kNumLenToPosStates - 1) << kNumPosSlotBits); - TREE_6_DECODE(prob, distance); - if (distance >= kStartPosModelIndex) - { - unsigned posSlot = (unsigned)distance; - unsigned numDirectBits = (unsigned)(((distance >> 1) - 1)); - distance = (2 | (distance & 1)); - if (posSlot < kEndPosModelIndex) - { - distance <<= numDirectBits; - prob = probs + SpecPos; - { - UInt32 m = 1; - distance++; - do - { - REV_BIT_VAR(prob, distance, m); - } - while (--numDirectBits); - distance -= m; - } - } - else - { - numDirectBits -= kNumAlignBits; - do - { - NORMALIZE - range >>= 1; - - { - UInt32 t; - code -= range; - t = (0 - ((UInt32)code >> 31)); /* (UInt32)((Int32)code >> 31) */ - distance = (distance << 1) + (t + 1); - code += range & t; - } - /* - distance <<= 1; - if (code >= range) - { - code -= range; - distance |= 1; - } - */ - } - while (--numDirectBits); - prob = probs + Align; - distance <<= kNumAlignBits; - { - unsigned i = 1; - REV_BIT_CONST(prob, i, 1); - REV_BIT_CONST(prob, i, 2); - REV_BIT_CONST(prob, i, 4); - REV_BIT_LAST (prob, i, 8); - distance |= i; - } - if (distance == (UInt32)0xFFFFFFFF) - { - len = kMatchSpecLenStart; - state -= kNumStates; - break; - } - } - } - - rep3 = rep2; - rep2 = rep1; - rep1 = rep0; - rep0 = distance + 1; - state = (state < kNumStates + kNumLitStates) ? kNumLitStates : kNumLitStates + 3; - if (distance >= (checkDicSize == 0 ? processedPos: checkDicSize)) - { - len += kMatchSpecLen_Error_Data + kMatchMinLen; - // len = kMatchSpecLen_Error_Data; - // len += kMatchMinLen; - break; - } - } - - len += kMatchMinLen; - - { - SizeT rem; - unsigned curLen; - SizeT pos; - - if ((rem = limit - dicPos) == 0) - { - /* - We stop decoding and return SZ_OK, and we can resume decoding later. - Any error conditions can be tested later in caller code. - For more strict mode we can stop decoding with error - // len += kMatchSpecLen_Error_Data; - */ - break; - } - - curLen = ((rem < len) ? (unsigned)rem : len); - pos = dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0); - - processedPos += (UInt32)curLen; - - len -= curLen; - if (curLen <= dicBufSize - pos) - { - Byte *dest = dic + dicPos; - ptrdiff_t src = (ptrdiff_t)pos - (ptrdiff_t)dicPos; - const Byte *lim = dest + curLen; - dicPos += (SizeT)curLen; - do - *(dest) = (Byte)*(dest + src); - while (++dest != lim); - } - else - { - do - { - dic[dicPos++] = dic[pos]; - if (++pos == dicBufSize) - pos = 0; - } - while (--curLen != 0); - } - } - } - } - while (dicPos < limit && buf < bufLimit); - - NORMALIZE; - - p->buf = buf; - p->range = range; - p->code = code; - p->remainLen = (UInt32)len; // & (kMatchSpecLen_Error_Data - 1); // we can write real length for error matches too. - p->dicPos = dicPos; - p->processedPos = processedPos; - p->reps[0] = rep0; - p->reps[1] = rep1; - p->reps[2] = rep2; - p->reps[3] = rep3; - p->state = (UInt32)state; - if (len >= kMatchSpecLen_Error_Data) - return SZ_ERROR_DATA; - return SZ_OK; -} -#endif - - - -static void MY_FAST_CALL LzmaDec_WriteRem(CLzmaDec *p, SizeT limit) -{ - unsigned len = (unsigned)p->remainLen; - if (len == 0 /* || len >= kMatchSpecLenStart */) - return; - { - SizeT dicPos = p->dicPos; - Byte *dic; - SizeT dicBufSize; - SizeT rep0; /* we use SizeT to avoid the BUG of VC14 for AMD64 */ - { - SizeT rem = limit - dicPos; - if (rem < len) - { - len = (unsigned)(rem); - if (len == 0) - return; - } - } - - if (p->checkDicSize == 0 && p->prop.dicSize - p->processedPos <= len) - p->checkDicSize = p->prop.dicSize; - - p->processedPos += (UInt32)len; - p->remainLen -= (UInt32)len; - dic = p->dic; - rep0 = p->reps[0]; - dicBufSize = p->dicBufSize; - do - { - dic[dicPos] = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)]; - dicPos++; - } - while (--len); - p->dicPos = dicPos; - } -} - - -/* -At staring of new stream we have one of the following symbols: - - Literal - is allowed - - Non-Rep-Match - is allowed only if it's end marker symbol - - Rep-Match - is not allowed -We use early check of (RangeCoder:Code) over kBadRepCode to simplify main decoding code -*/ - -#define kRange0 0xFFFFFFFF -#define kBound0 ((kRange0 >> kNumBitModelTotalBits) << (kNumBitModelTotalBits - 1)) -#define kBadRepCode (kBound0 + (((kRange0 - kBound0) >> kNumBitModelTotalBits) << (kNumBitModelTotalBits - 1))) -#if kBadRepCode != (0xC0000000 - 0x400) - #error Stop_Compiling_Bad_LZMA_Check -#endif - - -/* -LzmaDec_DecodeReal2(): - It calls LZMA_DECODE_REAL() and it adjusts limit according (p->checkDicSize). - -We correct (p->checkDicSize) after LZMA_DECODE_REAL() and in LzmaDec_WriteRem(), -and we support the following state of (p->checkDicSize): - if (total_processed < p->prop.dicSize) then - { - (total_processed == p->processedPos) - (p->checkDicSize == 0) - } - else - (p->checkDicSize == p->prop.dicSize) -*/ - -static int MY_FAST_CALL LzmaDec_DecodeReal2(CLzmaDec *p, SizeT limit, const Byte *bufLimit) -{ - if (p->checkDicSize == 0) - { - UInt32 rem = p->prop.dicSize - p->processedPos; - if (limit - p->dicPos > rem) - limit = p->dicPos + rem; - } - { - int res = LZMA_DECODE_REAL(p, limit, bufLimit); - if (p->checkDicSize == 0 && p->processedPos >= p->prop.dicSize) - p->checkDicSize = p->prop.dicSize; - return res; - } -} - - - -typedef enum -{ - DUMMY_INPUT_EOF, /* need more input data */ - DUMMY_LIT, - DUMMY_MATCH, - DUMMY_REP -} ELzmaDummy; - - -#define IS_DUMMY_END_MARKER_POSSIBLE(dummyRes) ((dummyRes) == DUMMY_MATCH) - -static ELzmaDummy LzmaDec_TryDummy(const CLzmaDec *p, const Byte *buf, const Byte **bufOut) -{ - UInt32 range = p->range; - UInt32 code = p->code; - const Byte *bufLimit = *bufOut; - const CLzmaProb *probs = GET_PROBS; - unsigned state = (unsigned)p->state; - ELzmaDummy res; - - for (;;) - { - const CLzmaProb *prob; - UInt32 bound; - unsigned ttt; - unsigned posState = CALC_POS_STATE(p->processedPos, ((unsigned)1 << p->prop.pb) - 1); - - prob = probs + IsMatch + COMBINED_PS_STATE; - IF_BIT_0_CHECK(prob) - { - UPDATE_0_CHECK - - prob = probs + Literal; - if (p->checkDicSize != 0 || p->processedPos != 0) - prob += ((UInt32)LZMA_LIT_SIZE * - ((((p->processedPos) & (((unsigned)1 << (p->prop.lp)) - 1)) << p->prop.lc) + - ((unsigned)p->dic[(p->dicPos == 0 ? p->dicBufSize : p->dicPos) - 1] >> (8 - p->prop.lc)))); - - if (state < kNumLitStates) - { - unsigned symbol = 1; - do { GET_BIT_CHECK(prob + symbol, symbol) } while (symbol < 0x100); - } - else - { - unsigned matchByte = p->dic[p->dicPos - p->reps[0] + - (p->dicPos < p->reps[0] ? p->dicBufSize : 0)]; - unsigned offs = 0x100; - unsigned symbol = 1; - do - { - unsigned bit; - const CLzmaProb *probLit; - matchByte += matchByte; - bit = offs; - offs &= matchByte; - probLit = prob + (offs + bit + symbol); - GET_BIT2_CHECK(probLit, symbol, offs ^= bit; , ; ) - } - while (symbol < 0x100); - } - res = DUMMY_LIT; - } - else - { - unsigned len; - UPDATE_1_CHECK; - - prob = probs + IsRep + state; - IF_BIT_0_CHECK(prob) - { - UPDATE_0_CHECK; - state = 0; - prob = probs + LenCoder; - res = DUMMY_MATCH; - } - else - { - UPDATE_1_CHECK; - res = DUMMY_REP; - prob = probs + IsRepG0 + state; - IF_BIT_0_CHECK(prob) - { - UPDATE_0_CHECK; - prob = probs + IsRep0Long + COMBINED_PS_STATE; - IF_BIT_0_CHECK(prob) - { - UPDATE_0_CHECK; - break; - } - else - { - UPDATE_1_CHECK; - } - } - else - { - UPDATE_1_CHECK; - prob = probs + IsRepG1 + state; - IF_BIT_0_CHECK(prob) - { - UPDATE_0_CHECK; - } - else - { - UPDATE_1_CHECK; - prob = probs + IsRepG2 + state; - IF_BIT_0_CHECK(prob) - { - UPDATE_0_CHECK; - } - else - { - UPDATE_1_CHECK; - } - } - } - state = kNumStates; - prob = probs + RepLenCoder; - } - { - unsigned limit, offset; - const CLzmaProb *probLen = prob + LenChoice; - IF_BIT_0_CHECK(probLen) - { - UPDATE_0_CHECK; - probLen = prob + LenLow + GET_LEN_STATE; - offset = 0; - limit = 1 << kLenNumLowBits; - } - else - { - UPDATE_1_CHECK; - probLen = prob + LenChoice2; - IF_BIT_0_CHECK(probLen) - { - UPDATE_0_CHECK; - probLen = prob + LenLow + GET_LEN_STATE + (1 << kLenNumLowBits); - offset = kLenNumLowSymbols; - limit = 1 << kLenNumLowBits; - } - else - { - UPDATE_1_CHECK; - probLen = prob + LenHigh; - offset = kLenNumLowSymbols * 2; - limit = 1 << kLenNumHighBits; - } - } - TREE_DECODE_CHECK(probLen, limit, len); - len += offset; - } - - if (state < 4) - { - unsigned posSlot; - prob = probs + PosSlot + - ((len < kNumLenToPosStates - 1 ? len : kNumLenToPosStates - 1) << - kNumPosSlotBits); - TREE_DECODE_CHECK(prob, 1 << kNumPosSlotBits, posSlot); - if (posSlot >= kStartPosModelIndex) - { - unsigned numDirectBits = ((posSlot >> 1) - 1); - - if (posSlot < kEndPosModelIndex) - { - prob = probs + SpecPos + ((2 | (posSlot & 1)) << numDirectBits); - } - else - { - numDirectBits -= kNumAlignBits; - do - { - NORMALIZE_CHECK - range >>= 1; - code -= range & (((code - range) >> 31) - 1); - /* if (code >= range) code -= range; */ - } - while (--numDirectBits); - prob = probs + Align; - numDirectBits = kNumAlignBits; - } - { - unsigned i = 1; - unsigned m = 1; - do - { - REV_BIT_CHECK(prob, i, m); - } - while (--numDirectBits); - } - } - } - } - break; - } - NORMALIZE_CHECK; - - *bufOut = buf; - return res; -} - -void LzmaDec_InitDicAndState(CLzmaDec *p, BoolInt initDic, BoolInt initState); -void LzmaDec_InitDicAndState(CLzmaDec *p, BoolInt initDic, BoolInt initState) -{ - p->remainLen = kMatchSpecLenStart + 1; - p->tempBufSize = 0; - - if (initDic) - { - p->processedPos = 0; - p->checkDicSize = 0; - p->remainLen = kMatchSpecLenStart + 2; - } - if (initState) - p->remainLen = kMatchSpecLenStart + 2; -} - -void LzmaDec_Init(CLzmaDec *p) -{ - p->dicPos = 0; - LzmaDec_InitDicAndState(p, True, True); -} - - -/* -LZMA supports optional end_marker. -So the decoder can lookahead for one additional LZMA-Symbol to check end_marker. -That additional LZMA-Symbol can require up to LZMA_REQUIRED_INPUT_MAX bytes in input stream. -When the decoder reaches dicLimit, it looks (finishMode) parameter: - if (finishMode == LZMA_FINISH_ANY), the decoder doesn't lookahead - if (finishMode != LZMA_FINISH_ANY), the decoder lookahead, if end_marker is possible for current position - -When the decoder lookahead, and the lookahead symbol is not end_marker, we have two ways: - 1) Strict mode (default) : the decoder returns SZ_ERROR_DATA. - 2) The relaxed mode (alternative mode) : we could return SZ_OK, and the caller - must check (status) value. The caller can show the error, - if the end of stream is expected, and the (status) is noit - LZMA_STATUS_FINISHED_WITH_MARK or LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK. -*/ - - -#define RETURN__NOT_FINISHED__FOR_FINISH \ - *status = LZMA_STATUS_NOT_FINISHED; \ - return SZ_ERROR_DATA; // for strict mode - // return SZ_OK; // for relaxed mode - - -SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit, const Byte *src, SizeT *srcLen, - ELzmaFinishMode finishMode, ELzmaStatus *status) -{ - SizeT inSize = *srcLen; - (*srcLen) = 0; - *status = LZMA_STATUS_NOT_SPECIFIED; - - if (p->remainLen > kMatchSpecLenStart) - { - if (p->remainLen > kMatchSpecLenStart + 2) - return p->remainLen == kMatchSpecLen_Error_Fail ? SZ_ERROR_FAIL : SZ_ERROR_DATA; - - for (; inSize > 0 && p->tempBufSize < RC_INIT_SIZE; (*srcLen)++, inSize--) - p->tempBuf[p->tempBufSize++] = *src++; - if (p->tempBufSize != 0 && p->tempBuf[0] != 0) - return SZ_ERROR_DATA; - if (p->tempBufSize < RC_INIT_SIZE) - { - *status = LZMA_STATUS_NEEDS_MORE_INPUT; - return SZ_OK; - } - p->code = - ((UInt32)p->tempBuf[1] << 24) - | ((UInt32)p->tempBuf[2] << 16) - | ((UInt32)p->tempBuf[3] << 8) - | ((UInt32)p->tempBuf[4]); - - if (p->checkDicSize == 0 - && p->processedPos == 0 - && p->code >= kBadRepCode) - return SZ_ERROR_DATA; - - p->range = 0xFFFFFFFF; - p->tempBufSize = 0; - - if (p->remainLen > kMatchSpecLenStart + 1) - { - SizeT numProbs = LzmaProps_GetNumProbs(&p->prop); - SizeT i; - CLzmaProb *probs = p->probs; - for (i = 0; i < numProbs; i++) - probs[i] = kBitModelTotal >> 1; - p->reps[0] = p->reps[1] = p->reps[2] = p->reps[3] = 1; - p->state = 0; - } - - p->remainLen = 0; - } - - for (;;) - { - if (p->remainLen == kMatchSpecLenStart) - { - if (p->code != 0) - return SZ_ERROR_DATA; - *status = LZMA_STATUS_FINISHED_WITH_MARK; - return SZ_OK; - } - - LzmaDec_WriteRem(p, dicLimit); - - { - // (p->remainLen == 0 || p->dicPos == dicLimit) - - int checkEndMarkNow = 0; - - if (p->dicPos >= dicLimit) - { - if (p->remainLen == 0 && p->code == 0) - { - *status = LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK; - return SZ_OK; - } - if (finishMode == LZMA_FINISH_ANY) - { - *status = LZMA_STATUS_NOT_FINISHED; - return SZ_OK; - } - if (p->remainLen != 0) - { - RETURN__NOT_FINISHED__FOR_FINISH; - } - checkEndMarkNow = 1; - } - - // (p->remainLen == 0) - - if (p->tempBufSize == 0) - { - const Byte *bufLimit; - int dummyProcessed = -1; - - if (inSize < LZMA_REQUIRED_INPUT_MAX || checkEndMarkNow) - { - const Byte *bufOut = src + inSize; - - ELzmaDummy dummyRes = LzmaDec_TryDummy(p, src, &bufOut); - - if (dummyRes == DUMMY_INPUT_EOF) - { - size_t i; - if (inSize >= LZMA_REQUIRED_INPUT_MAX) - break; - (*srcLen) += inSize; - p->tempBufSize = (unsigned)inSize; - for (i = 0; i < inSize; i++) - p->tempBuf[i] = src[i]; - *status = LZMA_STATUS_NEEDS_MORE_INPUT; - return SZ_OK; - } - - dummyProcessed = (int)(bufOut - src); - if ((unsigned)dummyProcessed > LZMA_REQUIRED_INPUT_MAX) - break; - - if (checkEndMarkNow && !IS_DUMMY_END_MARKER_POSSIBLE(dummyRes)) - { - unsigned i; - (*srcLen) += (unsigned)dummyProcessed; - p->tempBufSize = (unsigned)dummyProcessed; - for (i = 0; i < (unsigned)dummyProcessed; i++) - p->tempBuf[i] = src[i]; - // p->remainLen = kMatchSpecLen_Error_Data; - RETURN__NOT_FINISHED__FOR_FINISH; - } - - bufLimit = src; - // we will decode only one iteration - } - else - bufLimit = src + inSize - LZMA_REQUIRED_INPUT_MAX; - - p->buf = src; - - { - int res = LzmaDec_DecodeReal2(p, dicLimit, bufLimit); - - SizeT processed = (SizeT)(p->buf - src); - - if (dummyProcessed < 0) - { - if (processed > inSize) - break; - } - else if ((unsigned)dummyProcessed != processed) - break; - - src += processed; - inSize -= processed; - (*srcLen) += processed; - - if (res != SZ_OK) - { - p->remainLen = kMatchSpecLen_Error_Data; - return SZ_ERROR_DATA; - } - } - continue; - } - - { - // we have some data in (p->tempBuf) - // in strict mode: tempBufSize is not enough for one Symbol decoding. - // in relaxed mode: tempBufSize not larger than required for one Symbol decoding. - - unsigned rem = p->tempBufSize; - unsigned ahead = 0; - int dummyProcessed = -1; - - while (rem < LZMA_REQUIRED_INPUT_MAX && ahead < inSize) - p->tempBuf[rem++] = src[ahead++]; - - // ahead - the size of new data copied from (src) to (p->tempBuf) - // rem - the size of temp buffer including new data from (src) - - if (rem < LZMA_REQUIRED_INPUT_MAX || checkEndMarkNow) - { - const Byte *bufOut = p->tempBuf + rem; - - ELzmaDummy dummyRes = LzmaDec_TryDummy(p, p->tempBuf, &bufOut); - - if (dummyRes == DUMMY_INPUT_EOF) - { - if (rem >= LZMA_REQUIRED_INPUT_MAX) - break; - p->tempBufSize = rem; - (*srcLen) += (SizeT)ahead; - *status = LZMA_STATUS_NEEDS_MORE_INPUT; - return SZ_OK; - } - - dummyProcessed = (int)(bufOut - p->tempBuf); - - if ((unsigned)dummyProcessed < p->tempBufSize) - break; - - if (checkEndMarkNow && !IS_DUMMY_END_MARKER_POSSIBLE(dummyRes)) - { - (*srcLen) += (unsigned)dummyProcessed - p->tempBufSize; - p->tempBufSize = (unsigned)dummyProcessed; - // p->remainLen = kMatchSpecLen_Error_Data; - RETURN__NOT_FINISHED__FOR_FINISH; - } - } - - p->buf = p->tempBuf; - - { - // we decode one symbol from (p->tempBuf) here, so the (bufLimit) is equal to (p->buf) - int res = LzmaDec_DecodeReal2(p, dicLimit, p->buf); - - SizeT processed = (SizeT)(p->buf - p->tempBuf); - rem = p->tempBufSize; - - if (dummyProcessed < 0) - { - if (processed > LZMA_REQUIRED_INPUT_MAX) - break; - if (processed < rem) - break; - } - else if ((unsigned)dummyProcessed != processed) - break; - - processed -= rem; - - src += processed; - inSize -= processed; - (*srcLen) += processed; - p->tempBufSize = 0; - - if (res != SZ_OK) - { - p->remainLen = kMatchSpecLen_Error_Data; - return SZ_ERROR_DATA; - } - } - } - } - } - - /* Some unexpected error: internal error of code, memory corruption or hardware failure */ - p->remainLen = kMatchSpecLen_Error_Fail; - return SZ_ERROR_FAIL; -} - - - -SRes LzmaDec_DecodeToBuf(CLzmaDec *p, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status) -{ - SizeT outSize = *destLen; - SizeT inSize = *srcLen; - *srcLen = *destLen = 0; - for (;;) - { - SizeT inSizeCur = inSize, outSizeCur, dicPos; - ELzmaFinishMode curFinishMode; - SRes res; - if (p->dicPos == p->dicBufSize) - p->dicPos = 0; - dicPos = p->dicPos; - if (outSize > p->dicBufSize - dicPos) - { - outSizeCur = p->dicBufSize; - curFinishMode = LZMA_FINISH_ANY; - } - else - { - outSizeCur = dicPos + outSize; - curFinishMode = finishMode; - } - - res = LzmaDec_DecodeToDic(p, outSizeCur, src, &inSizeCur, curFinishMode, status); - src += inSizeCur; - inSize -= inSizeCur; - *srcLen += inSizeCur; - outSizeCur = p->dicPos - dicPos; - memcpy(dest, p->dic + dicPos, outSizeCur); - dest += outSizeCur; - outSize -= outSizeCur; - *destLen += outSizeCur; - if (res != 0) - return res; - if (outSizeCur == 0 || outSize == 0) - return SZ_OK; - } -} - -void LzmaDec_FreeProbs(CLzmaDec *p, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, p->probs); - p->probs = NULL; -} - -static void LzmaDec_FreeDict(CLzmaDec *p, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, p->dic); - p->dic = NULL; -} - -void LzmaDec_Free(CLzmaDec *p, ISzAllocPtr alloc) -{ - LzmaDec_FreeProbs(p, alloc); - LzmaDec_FreeDict(p, alloc); -} - -SRes LzmaProps_Decode(CLzmaProps *p, const Byte *data, unsigned size) -{ - UInt32 dicSize; - Byte d; - - if (size < LZMA_PROPS_SIZE) - return SZ_ERROR_UNSUPPORTED; - else - dicSize = data[1] | ((UInt32)data[2] << 8) | ((UInt32)data[3] << 16) | ((UInt32)data[4] << 24); - - if (dicSize < LZMA_DIC_MIN) - dicSize = LZMA_DIC_MIN; - p->dicSize = dicSize; - - d = data[0]; - if (d >= (9 * 5 * 5)) - return SZ_ERROR_UNSUPPORTED; - - p->lc = (Byte)(d % 9); - d /= 9; - p->pb = (Byte)(d / 5); - p->lp = (Byte)(d % 5); - - return SZ_OK; -} - -static SRes LzmaDec_AllocateProbs2(CLzmaDec *p, const CLzmaProps *propNew, ISzAllocPtr alloc) -{ - UInt32 numProbs = LzmaProps_GetNumProbs(propNew); - if (!p->probs || numProbs != p->numProbs) - { - LzmaDec_FreeProbs(p, alloc); - p->probs = (CLzmaProb *)ISzAlloc_Alloc(alloc, numProbs * sizeof(CLzmaProb)); - if (!p->probs) - return SZ_ERROR_MEM; - p->probs_1664 = p->probs + 1664; - p->numProbs = numProbs; - } - return SZ_OK; -} - -SRes LzmaDec_AllocateProbs(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc) -{ - CLzmaProps propNew; - RINOK(LzmaProps_Decode(&propNew, props, propsSize)); - RINOK(LzmaDec_AllocateProbs2(p, &propNew, alloc)); - p->prop = propNew; - return SZ_OK; -} - -SRes LzmaDec_Allocate(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc) -{ - CLzmaProps propNew; - SizeT dicBufSize; - RINOK(LzmaProps_Decode(&propNew, props, propsSize)); - RINOK(LzmaDec_AllocateProbs2(p, &propNew, alloc)); - - { - UInt32 dictSize = propNew.dicSize; - SizeT mask = ((UInt32)1 << 12) - 1; - if (dictSize >= ((UInt32)1 << 30)) mask = ((UInt32)1 << 22) - 1; - else if (dictSize >= ((UInt32)1 << 22)) mask = ((UInt32)1 << 20) - 1;; - dicBufSize = ((SizeT)dictSize + mask) & ~mask; - if (dicBufSize < dictSize) - dicBufSize = dictSize; - } - - if (!p->dic || dicBufSize != p->dicBufSize) - { - LzmaDec_FreeDict(p, alloc); - p->dic = (Byte *)ISzAlloc_Alloc(alloc, dicBufSize); - if (!p->dic) - { - LzmaDec_FreeProbs(p, alloc); - return SZ_ERROR_MEM; - } - } - p->dicBufSize = dicBufSize; - p->prop = propNew; - return SZ_OK; -} - -SRes LzmaDecode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, - const Byte *propData, unsigned propSize, ELzmaFinishMode finishMode, - ELzmaStatus *status, ISzAllocPtr alloc) -{ - CLzmaDec p; - SRes res; - SizeT outSize = *destLen, inSize = *srcLen; - *destLen = *srcLen = 0; - *status = LZMA_STATUS_NOT_SPECIFIED; - if (inSize < RC_INIT_SIZE) - return SZ_ERROR_INPUT_EOF; - LzmaDec_Construct(&p); - RINOK(LzmaDec_AllocateProbs(&p, propData, propSize, alloc)); - p.dic = dest; - p.dicBufSize = outSize; - LzmaDec_Init(&p); - *srcLen = inSize; - res = LzmaDec_DecodeToDic(&p, outSize, src, srcLen, finishMode, status); - *destLen = p.dicPos; - if (res == SZ_OK && *status == LZMA_STATUS_NEEDS_MORE_INPUT) - res = SZ_ERROR_INPUT_EOF; - LzmaDec_FreeProbs(&p, alloc); - return res; -} +/* LzmaDec.c -- LZMA Decoder +2021-04-01 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +/* #include "CpuArch.h" */ +#include "LzmaDec.h" + +#define kNumTopBits 24 +#define kTopValue ((UInt32)1 << kNumTopBits) + +#define kNumBitModelTotalBits 11 +#define kBitModelTotal (1 << kNumBitModelTotalBits) + +#define RC_INIT_SIZE 5 + +#ifndef _LZMA_DEC_OPT + +#define kNumMoveBits 5 +#define NORMALIZE if (range < kTopValue) { range <<= 8; code = (code << 8) | (*buf++); } + +#define IF_BIT_0(p) ttt = *(p); NORMALIZE; bound = (range >> kNumBitModelTotalBits) * (UInt32)ttt; if (code < bound) +#define UPDATE_0(p) range = bound; *(p) = (CLzmaProb)(ttt + ((kBitModelTotal - ttt) >> kNumMoveBits)); +#define UPDATE_1(p) range -= bound; code -= bound; *(p) = (CLzmaProb)(ttt - (ttt >> kNumMoveBits)); +#define GET_BIT2(p, i, A0, A1) IF_BIT_0(p) \ + { UPDATE_0(p); i = (i + i); A0; } else \ + { UPDATE_1(p); i = (i + i) + 1; A1; } + +#define TREE_GET_BIT(probs, i) { GET_BIT2(probs + i, i, ;, ;); } + +#define REV_BIT(p, i, A0, A1) IF_BIT_0(p + i) \ + { UPDATE_0(p + i); A0; } else \ + { UPDATE_1(p + i); A1; } +#define REV_BIT_VAR( p, i, m) REV_BIT(p, i, i += m; m += m, m += m; i += m; ) +#define REV_BIT_CONST(p, i, m) REV_BIT(p, i, i += m; , i += m * 2; ) +#define REV_BIT_LAST( p, i, m) REV_BIT(p, i, i -= m , ; ) + +#define TREE_DECODE(probs, limit, i) \ + { i = 1; do { TREE_GET_BIT(probs, i); } while (i < limit); i -= limit; } + +/* #define _LZMA_SIZE_OPT */ + +#ifdef _LZMA_SIZE_OPT +#define TREE_6_DECODE(probs, i) TREE_DECODE(probs, (1 << 6), i) +#else +#define TREE_6_DECODE(probs, i) \ + { i = 1; \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + i -= 0x40; } +#endif + +#define NORMAL_LITER_DEC TREE_GET_BIT(prob, symbol) +#define MATCHED_LITER_DEC \ + matchByte += matchByte; \ + bit = offs; \ + offs &= matchByte; \ + probLit = prob + (offs + bit + symbol); \ + GET_BIT2(probLit, symbol, offs ^= bit; , ;) + +#endif // _LZMA_DEC_OPT + + +#define NORMALIZE_CHECK if (range < kTopValue) { if (buf >= bufLimit) return DUMMY_INPUT_EOF; range <<= 8; code = (code << 8) | (*buf++); } + +#define IF_BIT_0_CHECK(p) ttt = *(p); NORMALIZE_CHECK; bound = (range >> kNumBitModelTotalBits) * (UInt32)ttt; if (code < bound) +#define UPDATE_0_CHECK range = bound; +#define UPDATE_1_CHECK range -= bound; code -= bound; +#define GET_BIT2_CHECK(p, i, A0, A1) IF_BIT_0_CHECK(p) \ + { UPDATE_0_CHECK; i = (i + i); A0; } else \ + { UPDATE_1_CHECK; i = (i + i) + 1; A1; } +#define GET_BIT_CHECK(p, i) GET_BIT2_CHECK(p, i, ; , ;) +#define TREE_DECODE_CHECK(probs, limit, i) \ + { i = 1; do { GET_BIT_CHECK(probs + i, i) } while (i < limit); i -= limit; } + + +#define REV_BIT_CHECK(p, i, m) IF_BIT_0_CHECK(p + i) \ + { UPDATE_0_CHECK; i += m; m += m; } else \ + { UPDATE_1_CHECK; m += m; i += m; } + + +#define kNumPosBitsMax 4 +#define kNumPosStatesMax (1 << kNumPosBitsMax) + +#define kLenNumLowBits 3 +#define kLenNumLowSymbols (1 << kLenNumLowBits) +#define kLenNumHighBits 8 +#define kLenNumHighSymbols (1 << kLenNumHighBits) + +#define LenLow 0 +#define LenHigh (LenLow + 2 * (kNumPosStatesMax << kLenNumLowBits)) +#define kNumLenProbs (LenHigh + kLenNumHighSymbols) + +#define LenChoice LenLow +#define LenChoice2 (LenLow + (1 << kLenNumLowBits)) + +#define kNumStates 12 +#define kNumStates2 16 +#define kNumLitStates 7 + +#define kStartPosModelIndex 4 +#define kEndPosModelIndex 14 +#define kNumFullDistances (1 << (kEndPosModelIndex >> 1)) + +#define kNumPosSlotBits 6 +#define kNumLenToPosStates 4 + +#define kNumAlignBits 4 +#define kAlignTableSize (1 << kNumAlignBits) + +#define kMatchMinLen 2 +#define kMatchSpecLenStart (kMatchMinLen + kLenNumLowSymbols * 2 + kLenNumHighSymbols) + +#define kMatchSpecLen_Error_Data (1 << 9) +#define kMatchSpecLen_Error_Fail (kMatchSpecLen_Error_Data - 1) + +/* External ASM code needs same CLzmaProb array layout. So don't change it. */ + +/* (probs_1664) is faster and better for code size at some platforms */ +/* +#ifdef MY_CPU_X86_OR_AMD64 +*/ +#define kStartOffset 1664 +#define GET_PROBS p->probs_1664 +/* +#define GET_PROBS p->probs + kStartOffset +#else +#define kStartOffset 0 +#define GET_PROBS p->probs +#endif +*/ + +#define SpecPos (-kStartOffset) +#define IsRep0Long (SpecPos + kNumFullDistances) +#define RepLenCoder (IsRep0Long + (kNumStates2 << kNumPosBitsMax)) +#define LenCoder (RepLenCoder + kNumLenProbs) +#define IsMatch (LenCoder + kNumLenProbs) +#define Align (IsMatch + (kNumStates2 << kNumPosBitsMax)) +#define IsRep (Align + kAlignTableSize) +#define IsRepG0 (IsRep + kNumStates) +#define IsRepG1 (IsRepG0 + kNumStates) +#define IsRepG2 (IsRepG1 + kNumStates) +#define PosSlot (IsRepG2 + kNumStates) +#define Literal (PosSlot + (kNumLenToPosStates << kNumPosSlotBits)) +#define NUM_BASE_PROBS (Literal + kStartOffset) + +#if Align != 0 && kStartOffset != 0 + #error Stop_Compiling_Bad_LZMA_kAlign +#endif + +#if NUM_BASE_PROBS != 1984 + #error Stop_Compiling_Bad_LZMA_PROBS +#endif + + +#define LZMA_LIT_SIZE 0x300 + +#define LzmaProps_GetNumProbs(p) (NUM_BASE_PROBS + ((UInt32)LZMA_LIT_SIZE << ((p)->lc + (p)->lp))) + + +#define CALC_POS_STATE(processedPos, pbMask) (((processedPos) & (pbMask)) << 4) +#define COMBINED_PS_STATE (posState + state) +#define GET_LEN_STATE (posState) + +#define LZMA_DIC_MIN (1 << 12) + +/* +p->remainLen : shows status of LZMA decoder: + < kMatchSpecLenStart : the number of bytes to be copied with (p->rep0) offset + = kMatchSpecLenStart : the LZMA stream was finished with end mark + = kMatchSpecLenStart + 1 : need init range coder + = kMatchSpecLenStart + 2 : need init range coder and state + = kMatchSpecLen_Error_Fail : Internal Code Failure + = kMatchSpecLen_Error_Data + [0 ... 273] : LZMA Data Error +*/ + +/* ---------- LZMA_DECODE_REAL ---------- */ +/* +LzmaDec_DecodeReal_3() can be implemented in external ASM file. +3 - is the code compatibility version of that function for check at link time. +*/ + +#define LZMA_DECODE_REAL LzmaDec_DecodeReal_3 + +/* +LZMA_DECODE_REAL() +In: + RangeCoder is normalized + if (p->dicPos == limit) + { + LzmaDec_TryDummy() was called before to exclude LITERAL and MATCH-REP cases. + So first symbol can be only MATCH-NON-REP. And if that MATCH-NON-REP symbol + is not END_OF_PAYALOAD_MARKER, then the function doesn't write any byte to dictionary, + the function returns SZ_OK, and the caller can use (p->remainLen) and (p->reps[0]) later. + } + +Processing: + The first LZMA symbol will be decoded in any case. + All main checks for limits are at the end of main loop, + It decodes additional LZMA-symbols while (p->buf < bufLimit && dicPos < limit), + RangeCoder is still without last normalization when (p->buf < bufLimit) is being checked. + But if (p->buf < bufLimit), the caller provided at least (LZMA_REQUIRED_INPUT_MAX + 1) bytes for + next iteration before limit (bufLimit + LZMA_REQUIRED_INPUT_MAX), + that is enough for worst case LZMA symbol with one additional RangeCoder normalization for one bit. + So that function never reads bufLimit [LZMA_REQUIRED_INPUT_MAX] byte. + +Out: + RangeCoder is normalized + Result: + SZ_OK - OK + p->remainLen: + < kMatchSpecLenStart : the number of bytes to be copied with (p->reps[0]) offset + = kMatchSpecLenStart : the LZMA stream was finished with end mark + + SZ_ERROR_DATA - error, when the MATCH-Symbol refers out of dictionary + p->remainLen : undefined + p->reps[*] : undefined +*/ + + +#ifdef _LZMA_DEC_OPT + +int MY_FAST_CALL LZMA_DECODE_REAL(CLzmaDec *p, SizeT limit, const Byte *bufLimit); + +#else + +static +int MY_FAST_CALL LZMA_DECODE_REAL(CLzmaDec *p, SizeT limit, const Byte *bufLimit) +{ + CLzmaProb *probs = GET_PROBS; + unsigned state = (unsigned)p->state; + UInt32 rep0 = p->reps[0], rep1 = p->reps[1], rep2 = p->reps[2], rep3 = p->reps[3]; + unsigned pbMask = ((unsigned)1 << (p->prop.pb)) - 1; + unsigned lc = p->prop.lc; + unsigned lpMask = ((unsigned)0x100 << p->prop.lp) - ((unsigned)0x100 >> lc); + + Byte *dic = p->dic; + SizeT dicBufSize = p->dicBufSize; + SizeT dicPos = p->dicPos; + + UInt32 processedPos = p->processedPos; + UInt32 checkDicSize = p->checkDicSize; + unsigned len = 0; + + const Byte *buf = p->buf; + UInt32 range = p->range; + UInt32 code = p->code; + + do + { + CLzmaProb *prob; + UInt32 bound; + unsigned ttt; + unsigned posState = CALC_POS_STATE(processedPos, pbMask); + + prob = probs + IsMatch + COMBINED_PS_STATE; + IF_BIT_0(prob) + { + unsigned symbol; + UPDATE_0(prob); + prob = probs + Literal; + if (processedPos != 0 || checkDicSize != 0) + prob += (UInt32)3 * ((((processedPos << 8) + dic[(dicPos == 0 ? dicBufSize : dicPos) - 1]) & lpMask) << lc); + processedPos++; + + if (state < kNumLitStates) + { + state -= (state < 4) ? state : 3; + symbol = 1; + #ifdef _LZMA_SIZE_OPT + do { NORMAL_LITER_DEC } while (symbol < 0x100); + #else + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + #endif + } + else + { + unsigned matchByte = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)]; + unsigned offs = 0x100; + state -= (state < 10) ? 3 : 6; + symbol = 1; + #ifdef _LZMA_SIZE_OPT + do + { + unsigned bit; + CLzmaProb *probLit; + MATCHED_LITER_DEC + } + while (symbol < 0x100); + #else + { + unsigned bit; + CLzmaProb *probLit; + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + } + #endif + } + + dic[dicPos++] = (Byte)symbol; + continue; + } + + { + UPDATE_1(prob); + prob = probs + IsRep + state; + IF_BIT_0(prob) + { + UPDATE_0(prob); + state += kNumStates; + prob = probs + LenCoder; + } + else + { + UPDATE_1(prob); + prob = probs + IsRepG0 + state; + IF_BIT_0(prob) + { + UPDATE_0(prob); + prob = probs + IsRep0Long + COMBINED_PS_STATE; + IF_BIT_0(prob) + { + UPDATE_0(prob); + + // that case was checked before with kBadRepCode + // if (checkDicSize == 0 && processedPos == 0) { len = kMatchSpecLen_Error_Data + 1; break; } + // The caller doesn't allow (dicPos == limit) case here + // so we don't need the following check: + // if (dicPos == limit) { state = state < kNumLitStates ? 9 : 11; len = 1; break; } + + dic[dicPos] = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)]; + dicPos++; + processedPos++; + state = state < kNumLitStates ? 9 : 11; + continue; + } + UPDATE_1(prob); + } + else + { + UInt32 distance; + UPDATE_1(prob); + prob = probs + IsRepG1 + state; + IF_BIT_0(prob) + { + UPDATE_0(prob); + distance = rep1; + } + else + { + UPDATE_1(prob); + prob = probs + IsRepG2 + state; + IF_BIT_0(prob) + { + UPDATE_0(prob); + distance = rep2; + } + else + { + UPDATE_1(prob); + distance = rep3; + rep3 = rep2; + } + rep2 = rep1; + } + rep1 = rep0; + rep0 = distance; + } + state = state < kNumLitStates ? 8 : 11; + prob = probs + RepLenCoder; + } + + #ifdef _LZMA_SIZE_OPT + { + unsigned lim, offset; + CLzmaProb *probLen = prob + LenChoice; + IF_BIT_0(probLen) + { + UPDATE_0(probLen); + probLen = prob + LenLow + GET_LEN_STATE; + offset = 0; + lim = (1 << kLenNumLowBits); + } + else + { + UPDATE_1(probLen); + probLen = prob + LenChoice2; + IF_BIT_0(probLen) + { + UPDATE_0(probLen); + probLen = prob + LenLow + GET_LEN_STATE + (1 << kLenNumLowBits); + offset = kLenNumLowSymbols; + lim = (1 << kLenNumLowBits); + } + else + { + UPDATE_1(probLen); + probLen = prob + LenHigh; + offset = kLenNumLowSymbols * 2; + lim = (1 << kLenNumHighBits); + } + } + TREE_DECODE(probLen, lim, len); + len += offset; + } + #else + { + CLzmaProb *probLen = prob + LenChoice; + IF_BIT_0(probLen) + { + UPDATE_0(probLen); + probLen = prob + LenLow + GET_LEN_STATE; + len = 1; + TREE_GET_BIT(probLen, len); + TREE_GET_BIT(probLen, len); + TREE_GET_BIT(probLen, len); + len -= 8; + } + else + { + UPDATE_1(probLen); + probLen = prob + LenChoice2; + IF_BIT_0(probLen) + { + UPDATE_0(probLen); + probLen = prob + LenLow + GET_LEN_STATE + (1 << kLenNumLowBits); + len = 1; + TREE_GET_BIT(probLen, len); + TREE_GET_BIT(probLen, len); + TREE_GET_BIT(probLen, len); + } + else + { + UPDATE_1(probLen); + probLen = prob + LenHigh; + TREE_DECODE(probLen, (1 << kLenNumHighBits), len); + len += kLenNumLowSymbols * 2; + } + } + } + #endif + + if (state >= kNumStates) + { + UInt32 distance; + prob = probs + PosSlot + + ((len < kNumLenToPosStates ? len : kNumLenToPosStates - 1) << kNumPosSlotBits); + TREE_6_DECODE(prob, distance); + if (distance >= kStartPosModelIndex) + { + unsigned posSlot = (unsigned)distance; + unsigned numDirectBits = (unsigned)(((distance >> 1) - 1)); + distance = (2 | (distance & 1)); + if (posSlot < kEndPosModelIndex) + { + distance <<= numDirectBits; + prob = probs + SpecPos; + { + UInt32 m = 1; + distance++; + do + { + REV_BIT_VAR(prob, distance, m); + } + while (--numDirectBits); + distance -= m; + } + } + else + { + numDirectBits -= kNumAlignBits; + do + { + NORMALIZE + range >>= 1; + + { + UInt32 t; + code -= range; + t = (0 - ((UInt32)code >> 31)); /* (UInt32)((Int32)code >> 31) */ + distance = (distance << 1) + (t + 1); + code += range & t; + } + /* + distance <<= 1; + if (code >= range) + { + code -= range; + distance |= 1; + } + */ + } + while (--numDirectBits); + prob = probs + Align; + distance <<= kNumAlignBits; + { + unsigned i = 1; + REV_BIT_CONST(prob, i, 1); + REV_BIT_CONST(prob, i, 2); + REV_BIT_CONST(prob, i, 4); + REV_BIT_LAST (prob, i, 8); + distance |= i; + } + if (distance == (UInt32)0xFFFFFFFF) + { + len = kMatchSpecLenStart; + state -= kNumStates; + break; + } + } + } + + rep3 = rep2; + rep2 = rep1; + rep1 = rep0; + rep0 = distance + 1; + state = (state < kNumStates + kNumLitStates) ? kNumLitStates : kNumLitStates + 3; + if (distance >= (checkDicSize == 0 ? processedPos: checkDicSize)) + { + len += kMatchSpecLen_Error_Data + kMatchMinLen; + // len = kMatchSpecLen_Error_Data; + // len += kMatchMinLen; + break; + } + } + + len += kMatchMinLen; + + { + SizeT rem; + unsigned curLen; + SizeT pos; + + if ((rem = limit - dicPos) == 0) + { + /* + We stop decoding and return SZ_OK, and we can resume decoding later. + Any error conditions can be tested later in caller code. + For more strict mode we can stop decoding with error + // len += kMatchSpecLen_Error_Data; + */ + break; + } + + curLen = ((rem < len) ? (unsigned)rem : len); + pos = dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0); + + processedPos += (UInt32)curLen; + + len -= curLen; + if (curLen <= dicBufSize - pos) + { + Byte *dest = dic + dicPos; + ptrdiff_t src = (ptrdiff_t)pos - (ptrdiff_t)dicPos; + const Byte *lim = dest + curLen; + dicPos += (SizeT)curLen; + do + *(dest) = (Byte)*(dest + src); + while (++dest != lim); + } + else + { + do + { + dic[dicPos++] = dic[pos]; + if (++pos == dicBufSize) + pos = 0; + } + while (--curLen != 0); + } + } + } + } + while (dicPos < limit && buf < bufLimit); + + NORMALIZE; + + p->buf = buf; + p->range = range; + p->code = code; + p->remainLen = (UInt32)len; // & (kMatchSpecLen_Error_Data - 1); // we can write real length for error matches too. + p->dicPos = dicPos; + p->processedPos = processedPos; + p->reps[0] = rep0; + p->reps[1] = rep1; + p->reps[2] = rep2; + p->reps[3] = rep3; + p->state = (UInt32)state; + if (len >= kMatchSpecLen_Error_Data) + return SZ_ERROR_DATA; + return SZ_OK; +} +#endif + + + +static void MY_FAST_CALL LzmaDec_WriteRem(CLzmaDec *p, SizeT limit) +{ + unsigned len = (unsigned)p->remainLen; + if (len == 0 /* || len >= kMatchSpecLenStart */) + return; + { + SizeT dicPos = p->dicPos; + Byte *dic; + SizeT dicBufSize; + SizeT rep0; /* we use SizeT to avoid the BUG of VC14 for AMD64 */ + { + SizeT rem = limit - dicPos; + if (rem < len) + { + len = (unsigned)(rem); + if (len == 0) + return; + } + } + + if (p->checkDicSize == 0 && p->prop.dicSize - p->processedPos <= len) + p->checkDicSize = p->prop.dicSize; + + p->processedPos += (UInt32)len; + p->remainLen -= (UInt32)len; + dic = p->dic; + rep0 = p->reps[0]; + dicBufSize = p->dicBufSize; + do + { + dic[dicPos] = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)]; + dicPos++; + } + while (--len); + p->dicPos = dicPos; + } +} + + +/* +At staring of new stream we have one of the following symbols: + - Literal - is allowed + - Non-Rep-Match - is allowed only if it's end marker symbol + - Rep-Match - is not allowed +We use early check of (RangeCoder:Code) over kBadRepCode to simplify main decoding code +*/ + +#define kRange0 0xFFFFFFFF +#define kBound0 ((kRange0 >> kNumBitModelTotalBits) << (kNumBitModelTotalBits - 1)) +#define kBadRepCode (kBound0 + (((kRange0 - kBound0) >> kNumBitModelTotalBits) << (kNumBitModelTotalBits - 1))) +#if kBadRepCode != (0xC0000000 - 0x400) + #error Stop_Compiling_Bad_LZMA_Check +#endif + + +/* +LzmaDec_DecodeReal2(): + It calls LZMA_DECODE_REAL() and it adjusts limit according (p->checkDicSize). + +We correct (p->checkDicSize) after LZMA_DECODE_REAL() and in LzmaDec_WriteRem(), +and we support the following state of (p->checkDicSize): + if (total_processed < p->prop.dicSize) then + { + (total_processed == p->processedPos) + (p->checkDicSize == 0) + } + else + (p->checkDicSize == p->prop.dicSize) +*/ + +static int MY_FAST_CALL LzmaDec_DecodeReal2(CLzmaDec *p, SizeT limit, const Byte *bufLimit) +{ + if (p->checkDicSize == 0) + { + UInt32 rem = p->prop.dicSize - p->processedPos; + if (limit - p->dicPos > rem) + limit = p->dicPos + rem; + } + { + int res = LZMA_DECODE_REAL(p, limit, bufLimit); + if (p->checkDicSize == 0 && p->processedPos >= p->prop.dicSize) + p->checkDicSize = p->prop.dicSize; + return res; + } +} + + + +typedef enum +{ + DUMMY_INPUT_EOF, /* need more input data */ + DUMMY_LIT, + DUMMY_MATCH, + DUMMY_REP +} ELzmaDummy; + + +#define IS_DUMMY_END_MARKER_POSSIBLE(dummyRes) ((dummyRes) == DUMMY_MATCH) + +static ELzmaDummy LzmaDec_TryDummy(const CLzmaDec *p, const Byte *buf, const Byte **bufOut) +{ + UInt32 range = p->range; + UInt32 code = p->code; + const Byte *bufLimit = *bufOut; + const CLzmaProb *probs = GET_PROBS; + unsigned state = (unsigned)p->state; + ELzmaDummy res; + + for (;;) + { + const CLzmaProb *prob; + UInt32 bound; + unsigned ttt; + unsigned posState = CALC_POS_STATE(p->processedPos, ((unsigned)1 << p->prop.pb) - 1); + + prob = probs + IsMatch + COMBINED_PS_STATE; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK + + prob = probs + Literal; + if (p->checkDicSize != 0 || p->processedPos != 0) + prob += ((UInt32)LZMA_LIT_SIZE * + ((((p->processedPos) & (((unsigned)1 << (p->prop.lp)) - 1)) << p->prop.lc) + + ((unsigned)p->dic[(p->dicPos == 0 ? p->dicBufSize : p->dicPos) - 1] >> (8 - p->prop.lc)))); + + if (state < kNumLitStates) + { + unsigned symbol = 1; + do { GET_BIT_CHECK(prob + symbol, symbol) } while (symbol < 0x100); + } + else + { + unsigned matchByte = p->dic[p->dicPos - p->reps[0] + + (p->dicPos < p->reps[0] ? p->dicBufSize : 0)]; + unsigned offs = 0x100; + unsigned symbol = 1; + do + { + unsigned bit; + const CLzmaProb *probLit; + matchByte += matchByte; + bit = offs; + offs &= matchByte; + probLit = prob + (offs + bit + symbol); + GET_BIT2_CHECK(probLit, symbol, offs ^= bit; , ; ) + } + while (symbol < 0x100); + } + res = DUMMY_LIT; + } + else + { + unsigned len; + UPDATE_1_CHECK; + + prob = probs + IsRep + state; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + state = 0; + prob = probs + LenCoder; + res = DUMMY_MATCH; + } + else + { + UPDATE_1_CHECK; + res = DUMMY_REP; + prob = probs + IsRepG0 + state; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + prob = probs + IsRep0Long + COMBINED_PS_STATE; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + break; + } + else + { + UPDATE_1_CHECK; + } + } + else + { + UPDATE_1_CHECK; + prob = probs + IsRepG1 + state; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + } + else + { + UPDATE_1_CHECK; + prob = probs + IsRepG2 + state; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + } + else + { + UPDATE_1_CHECK; + } + } + } + state = kNumStates; + prob = probs + RepLenCoder; + } + { + unsigned limit, offset; + const CLzmaProb *probLen = prob + LenChoice; + IF_BIT_0_CHECK(probLen) + { + UPDATE_0_CHECK; + probLen = prob + LenLow + GET_LEN_STATE; + offset = 0; + limit = 1 << kLenNumLowBits; + } + else + { + UPDATE_1_CHECK; + probLen = prob + LenChoice2; + IF_BIT_0_CHECK(probLen) + { + UPDATE_0_CHECK; + probLen = prob + LenLow + GET_LEN_STATE + (1 << kLenNumLowBits); + offset = kLenNumLowSymbols; + limit = 1 << kLenNumLowBits; + } + else + { + UPDATE_1_CHECK; + probLen = prob + LenHigh; + offset = kLenNumLowSymbols * 2; + limit = 1 << kLenNumHighBits; + } + } + TREE_DECODE_CHECK(probLen, limit, len); + len += offset; + } + + if (state < 4) + { + unsigned posSlot; + prob = probs + PosSlot + + ((len < kNumLenToPosStates - 1 ? len : kNumLenToPosStates - 1) << + kNumPosSlotBits); + TREE_DECODE_CHECK(prob, 1 << kNumPosSlotBits, posSlot); + if (posSlot >= kStartPosModelIndex) + { + unsigned numDirectBits = ((posSlot >> 1) - 1); + + if (posSlot < kEndPosModelIndex) + { + prob = probs + SpecPos + ((2 | (posSlot & 1)) << numDirectBits); + } + else + { + numDirectBits -= kNumAlignBits; + do + { + NORMALIZE_CHECK + range >>= 1; + code -= range & (((code - range) >> 31) - 1); + /* if (code >= range) code -= range; */ + } + while (--numDirectBits); + prob = probs + Align; + numDirectBits = kNumAlignBits; + } + { + unsigned i = 1; + unsigned m = 1; + do + { + REV_BIT_CHECK(prob, i, m); + } + while (--numDirectBits); + } + } + } + } + break; + } + NORMALIZE_CHECK; + + *bufOut = buf; + return res; +} + +void LzmaDec_InitDicAndState(CLzmaDec *p, BoolInt initDic, BoolInt initState); +void LzmaDec_InitDicAndState(CLzmaDec *p, BoolInt initDic, BoolInt initState) +{ + p->remainLen = kMatchSpecLenStart + 1; + p->tempBufSize = 0; + + if (initDic) + { + p->processedPos = 0; + p->checkDicSize = 0; + p->remainLen = kMatchSpecLenStart + 2; + } + if (initState) + p->remainLen = kMatchSpecLenStart + 2; +} + +void LzmaDec_Init(CLzmaDec *p) +{ + p->dicPos = 0; + LzmaDec_InitDicAndState(p, True, True); +} + + +/* +LZMA supports optional end_marker. +So the decoder can lookahead for one additional LZMA-Symbol to check end_marker. +That additional LZMA-Symbol can require up to LZMA_REQUIRED_INPUT_MAX bytes in input stream. +When the decoder reaches dicLimit, it looks (finishMode) parameter: + if (finishMode == LZMA_FINISH_ANY), the decoder doesn't lookahead + if (finishMode != LZMA_FINISH_ANY), the decoder lookahead, if end_marker is possible for current position + +When the decoder lookahead, and the lookahead symbol is not end_marker, we have two ways: + 1) Strict mode (default) : the decoder returns SZ_ERROR_DATA. + 2) The relaxed mode (alternative mode) : we could return SZ_OK, and the caller + must check (status) value. The caller can show the error, + if the end of stream is expected, and the (status) is noit + LZMA_STATUS_FINISHED_WITH_MARK or LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK. +*/ + + +#define RETURN__NOT_FINISHED__FOR_FINISH \ + *status = LZMA_STATUS_NOT_FINISHED; \ + return SZ_ERROR_DATA; // for strict mode + // return SZ_OK; // for relaxed mode + + +SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit, const Byte *src, SizeT *srcLen, + ELzmaFinishMode finishMode, ELzmaStatus *status) +{ + SizeT inSize = *srcLen; + (*srcLen) = 0; + *status = LZMA_STATUS_NOT_SPECIFIED; + + if (p->remainLen > kMatchSpecLenStart) + { + if (p->remainLen > kMatchSpecLenStart + 2) + return p->remainLen == kMatchSpecLen_Error_Fail ? SZ_ERROR_FAIL : SZ_ERROR_DATA; + + for (; inSize > 0 && p->tempBufSize < RC_INIT_SIZE; (*srcLen)++, inSize--) + p->tempBuf[p->tempBufSize++] = *src++; + if (p->tempBufSize != 0 && p->tempBuf[0] != 0) + return SZ_ERROR_DATA; + if (p->tempBufSize < RC_INIT_SIZE) + { + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + p->code = + ((UInt32)p->tempBuf[1] << 24) + | ((UInt32)p->tempBuf[2] << 16) + | ((UInt32)p->tempBuf[3] << 8) + | ((UInt32)p->tempBuf[4]); + + if (p->checkDicSize == 0 + && p->processedPos == 0 + && p->code >= kBadRepCode) + return SZ_ERROR_DATA; + + p->range = 0xFFFFFFFF; + p->tempBufSize = 0; + + if (p->remainLen > kMatchSpecLenStart + 1) + { + SizeT numProbs = LzmaProps_GetNumProbs(&p->prop); + SizeT i; + CLzmaProb *probs = p->probs; + for (i = 0; i < numProbs; i++) + probs[i] = kBitModelTotal >> 1; + p->reps[0] = p->reps[1] = p->reps[2] = p->reps[3] = 1; + p->state = 0; + } + + p->remainLen = 0; + } + + for (;;) + { + if (p->remainLen == kMatchSpecLenStart) + { + if (p->code != 0) + return SZ_ERROR_DATA; + *status = LZMA_STATUS_FINISHED_WITH_MARK; + return SZ_OK; + } + + LzmaDec_WriteRem(p, dicLimit); + + { + // (p->remainLen == 0 || p->dicPos == dicLimit) + + int checkEndMarkNow = 0; + + if (p->dicPos >= dicLimit) + { + if (p->remainLen == 0 && p->code == 0) + { + *status = LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK; + return SZ_OK; + } + if (finishMode == LZMA_FINISH_ANY) + { + *status = LZMA_STATUS_NOT_FINISHED; + return SZ_OK; + } + if (p->remainLen != 0) + { + RETURN__NOT_FINISHED__FOR_FINISH; + } + checkEndMarkNow = 1; + } + + // (p->remainLen == 0) + + if (p->tempBufSize == 0) + { + const Byte *bufLimit; + int dummyProcessed = -1; + + if (inSize < LZMA_REQUIRED_INPUT_MAX || checkEndMarkNow) + { + const Byte *bufOut = src + inSize; + + ELzmaDummy dummyRes = LzmaDec_TryDummy(p, src, &bufOut); + + if (dummyRes == DUMMY_INPUT_EOF) + { + size_t i; + if (inSize >= LZMA_REQUIRED_INPUT_MAX) + break; + (*srcLen) += inSize; + p->tempBufSize = (unsigned)inSize; + for (i = 0; i < inSize; i++) + p->tempBuf[i] = src[i]; + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + + dummyProcessed = (int)(bufOut - src); + if ((unsigned)dummyProcessed > LZMA_REQUIRED_INPUT_MAX) + break; + + if (checkEndMarkNow && !IS_DUMMY_END_MARKER_POSSIBLE(dummyRes)) + { + unsigned i; + (*srcLen) += (unsigned)dummyProcessed; + p->tempBufSize = (unsigned)dummyProcessed; + for (i = 0; i < (unsigned)dummyProcessed; i++) + p->tempBuf[i] = src[i]; + // p->remainLen = kMatchSpecLen_Error_Data; + RETURN__NOT_FINISHED__FOR_FINISH; + } + + bufLimit = src; + // we will decode only one iteration + } + else + bufLimit = src + inSize - LZMA_REQUIRED_INPUT_MAX; + + p->buf = src; + + { + int res = LzmaDec_DecodeReal2(p, dicLimit, bufLimit); + + SizeT processed = (SizeT)(p->buf - src); + + if (dummyProcessed < 0) + { + if (processed > inSize) + break; + } + else if ((unsigned)dummyProcessed != processed) + break; + + src += processed; + inSize -= processed; + (*srcLen) += processed; + + if (res != SZ_OK) + { + p->remainLen = kMatchSpecLen_Error_Data; + return SZ_ERROR_DATA; + } + } + continue; + } + + { + // we have some data in (p->tempBuf) + // in strict mode: tempBufSize is not enough for one Symbol decoding. + // in relaxed mode: tempBufSize not larger than required for one Symbol decoding. + + unsigned rem = p->tempBufSize; + unsigned ahead = 0; + int dummyProcessed = -1; + + while (rem < LZMA_REQUIRED_INPUT_MAX && ahead < inSize) + p->tempBuf[rem++] = src[ahead++]; + + // ahead - the size of new data copied from (src) to (p->tempBuf) + // rem - the size of temp buffer including new data from (src) + + if (rem < LZMA_REQUIRED_INPUT_MAX || checkEndMarkNow) + { + const Byte *bufOut = p->tempBuf + rem; + + ELzmaDummy dummyRes = LzmaDec_TryDummy(p, p->tempBuf, &bufOut); + + if (dummyRes == DUMMY_INPUT_EOF) + { + if (rem >= LZMA_REQUIRED_INPUT_MAX) + break; + p->tempBufSize = rem; + (*srcLen) += (SizeT)ahead; + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + + dummyProcessed = (int)(bufOut - p->tempBuf); + + if ((unsigned)dummyProcessed < p->tempBufSize) + break; + + if (checkEndMarkNow && !IS_DUMMY_END_MARKER_POSSIBLE(dummyRes)) + { + (*srcLen) += (unsigned)dummyProcessed - p->tempBufSize; + p->tempBufSize = (unsigned)dummyProcessed; + // p->remainLen = kMatchSpecLen_Error_Data; + RETURN__NOT_FINISHED__FOR_FINISH; + } + } + + p->buf = p->tempBuf; + + { + // we decode one symbol from (p->tempBuf) here, so the (bufLimit) is equal to (p->buf) + int res = LzmaDec_DecodeReal2(p, dicLimit, p->buf); + + SizeT processed = (SizeT)(p->buf - p->tempBuf); + rem = p->tempBufSize; + + if (dummyProcessed < 0) + { + if (processed > LZMA_REQUIRED_INPUT_MAX) + break; + if (processed < rem) + break; + } + else if ((unsigned)dummyProcessed != processed) + break; + + processed -= rem; + + src += processed; + inSize -= processed; + (*srcLen) += processed; + p->tempBufSize = 0; + + if (res != SZ_OK) + { + p->remainLen = kMatchSpecLen_Error_Data; + return SZ_ERROR_DATA; + } + } + } + } + } + + /* Some unexpected error: internal error of code, memory corruption or hardware failure */ + p->remainLen = kMatchSpecLen_Error_Fail; + return SZ_ERROR_FAIL; +} + + + +SRes LzmaDec_DecodeToBuf(CLzmaDec *p, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status) +{ + SizeT outSize = *destLen; + SizeT inSize = *srcLen; + *srcLen = *destLen = 0; + for (;;) + { + SizeT inSizeCur = inSize, outSizeCur, dicPos; + ELzmaFinishMode curFinishMode; + SRes res; + if (p->dicPos == p->dicBufSize) + p->dicPos = 0; + dicPos = p->dicPos; + if (outSize > p->dicBufSize - dicPos) + { + outSizeCur = p->dicBufSize; + curFinishMode = LZMA_FINISH_ANY; + } + else + { + outSizeCur = dicPos + outSize; + curFinishMode = finishMode; + } + + res = LzmaDec_DecodeToDic(p, outSizeCur, src, &inSizeCur, curFinishMode, status); + src += inSizeCur; + inSize -= inSizeCur; + *srcLen += inSizeCur; + outSizeCur = p->dicPos - dicPos; + memcpy(dest, p->dic + dicPos, outSizeCur); + dest += outSizeCur; + outSize -= outSizeCur; + *destLen += outSizeCur; + if (res != 0) + return res; + if (outSizeCur == 0 || outSize == 0) + return SZ_OK; + } +} + +void LzmaDec_FreeProbs(CLzmaDec *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->probs); + p->probs = NULL; +} + +static void LzmaDec_FreeDict(CLzmaDec *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->dic); + p->dic = NULL; +} + +void LzmaDec_Free(CLzmaDec *p, ISzAllocPtr alloc) +{ + LzmaDec_FreeProbs(p, alloc); + LzmaDec_FreeDict(p, alloc); +} + +SRes LzmaProps_Decode(CLzmaProps *p, const Byte *data, unsigned size) +{ + UInt32 dicSize; + Byte d; + + if (size < LZMA_PROPS_SIZE) + return SZ_ERROR_UNSUPPORTED; + else + dicSize = data[1] | ((UInt32)data[2] << 8) | ((UInt32)data[3] << 16) | ((UInt32)data[4] << 24); + + if (dicSize < LZMA_DIC_MIN) + dicSize = LZMA_DIC_MIN; + p->dicSize = dicSize; + + d = data[0]; + if (d >= (9 * 5 * 5)) + return SZ_ERROR_UNSUPPORTED; + + p->lc = (Byte)(d % 9); + d /= 9; + p->pb = (Byte)(d / 5); + p->lp = (Byte)(d % 5); + + return SZ_OK; +} + +static SRes LzmaDec_AllocateProbs2(CLzmaDec *p, const CLzmaProps *propNew, ISzAllocPtr alloc) +{ + UInt32 numProbs = LzmaProps_GetNumProbs(propNew); + if (!p->probs || numProbs != p->numProbs) + { + LzmaDec_FreeProbs(p, alloc); + p->probs = (CLzmaProb *)ISzAlloc_Alloc(alloc, numProbs * sizeof(CLzmaProb)); + if (!p->probs) + return SZ_ERROR_MEM; + p->probs_1664 = p->probs + 1664; + p->numProbs = numProbs; + } + return SZ_OK; +} + +SRes LzmaDec_AllocateProbs(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc) +{ + CLzmaProps propNew; + RINOK(LzmaProps_Decode(&propNew, props, propsSize)); + RINOK(LzmaDec_AllocateProbs2(p, &propNew, alloc)); + p->prop = propNew; + return SZ_OK; +} + +SRes LzmaDec_Allocate(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc) +{ + CLzmaProps propNew; + SizeT dicBufSize; + RINOK(LzmaProps_Decode(&propNew, props, propsSize)); + RINOK(LzmaDec_AllocateProbs2(p, &propNew, alloc)); + + { + UInt32 dictSize = propNew.dicSize; + SizeT mask = ((UInt32)1 << 12) - 1; + if (dictSize >= ((UInt32)1 << 30)) mask = ((UInt32)1 << 22) - 1; + else if (dictSize >= ((UInt32)1 << 22)) mask = ((UInt32)1 << 20) - 1;; + dicBufSize = ((SizeT)dictSize + mask) & ~mask; + if (dicBufSize < dictSize) + dicBufSize = dictSize; + } + + if (!p->dic || dicBufSize != p->dicBufSize) + { + LzmaDec_FreeDict(p, alloc); + p->dic = (Byte *)ISzAlloc_Alloc(alloc, dicBufSize); + if (!p->dic) + { + LzmaDec_FreeProbs(p, alloc); + return SZ_ERROR_MEM; + } + } + p->dicBufSize = dicBufSize; + p->prop = propNew; + return SZ_OK; +} + +SRes LzmaDecode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + const Byte *propData, unsigned propSize, ELzmaFinishMode finishMode, + ELzmaStatus *status, ISzAllocPtr alloc) +{ + CLzmaDec p; + SRes res; + SizeT outSize = *destLen, inSize = *srcLen; + *destLen = *srcLen = 0; + *status = LZMA_STATUS_NOT_SPECIFIED; + if (inSize < RC_INIT_SIZE) + return SZ_ERROR_INPUT_EOF; + LzmaDec_Construct(&p); + RINOK(LzmaDec_AllocateProbs(&p, propData, propSize, alloc)); + p.dic = dest; + p.dicBufSize = outSize; + LzmaDec_Init(&p); + *srcLen = inSize; + res = LzmaDec_DecodeToDic(&p, outSize, src, srcLen, finishMode, status); + *destLen = p.dicPos; + if (res == SZ_OK && *status == LZMA_STATUS_NEEDS_MORE_INPUT) + res = SZ_ERROR_INPUT_EOF; + LzmaDec_FreeProbs(&p, alloc); + return res; +} diff --git a/deps/LZMA-SDK/C/LzmaDec.h b/deps/LZMA-SDK/C/LzmaDec.h index 6194b7d12..6f1296250 100644 --- a/deps/LZMA-SDK/C/LzmaDec.h +++ b/deps/LZMA-SDK/C/LzmaDec.h @@ -1,236 +1,236 @@ -/* LzmaDec.h -- LZMA Decoder -2020-03-19 : Igor Pavlov : Public domain */ - -#ifndef __LZMA_DEC_H -#define __LZMA_DEC_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -/* #define _LZMA_PROB32 */ -/* _LZMA_PROB32 can increase the speed on some CPUs, - but memory usage for CLzmaDec::probs will be doubled in that case */ - -typedef -#ifdef _LZMA_PROB32 - UInt32 -#else - UInt16 -#endif - CLzmaProb; - - -/* ---------- LZMA Properties ---------- */ - -#define LZMA_PROPS_SIZE 5 - -typedef struct _CLzmaProps -{ - Byte lc; - Byte lp; - Byte pb; - Byte _pad_; - UInt32 dicSize; -} CLzmaProps; - -/* LzmaProps_Decode - decodes properties -Returns: - SZ_OK - SZ_ERROR_UNSUPPORTED - Unsupported properties -*/ - -SRes LzmaProps_Decode(CLzmaProps *p, const Byte *data, unsigned size); - - -/* ---------- LZMA Decoder state ---------- */ - -/* LZMA_REQUIRED_INPUT_MAX = number of required input bytes for worst case. - Num bits = log2((2^11 / 31) ^ 22) + 26 < 134 + 26 = 160; */ - -#define LZMA_REQUIRED_INPUT_MAX 20 - -typedef struct -{ - /* Don't change this structure. ASM code can use it. */ - CLzmaProps prop; - CLzmaProb *probs; - CLzmaProb *probs_1664; - Byte *dic; - SizeT dicBufSize; - SizeT dicPos; - const Byte *buf; - UInt32 range; - UInt32 code; - UInt32 processedPos; - UInt32 checkDicSize; - UInt32 reps[4]; - UInt32 state; - UInt32 remainLen; - - UInt32 numProbs; - unsigned tempBufSize; - Byte tempBuf[LZMA_REQUIRED_INPUT_MAX]; -} CLzmaDec; - -#define LzmaDec_Construct(p) { (p)->dic = NULL; (p)->probs = NULL; } - -void LzmaDec_Init(CLzmaDec *p); - -/* There are two types of LZMA streams: - - Stream with end mark. That end mark adds about 6 bytes to compressed size. - - Stream without end mark. You must know exact uncompressed size to decompress such stream. */ - -typedef enum -{ - LZMA_FINISH_ANY, /* finish at any point */ - LZMA_FINISH_END /* block must be finished at the end */ -} ELzmaFinishMode; - -/* ELzmaFinishMode has meaning only if the decoding reaches output limit !!! - - You must use LZMA_FINISH_END, when you know that current output buffer - covers last bytes of block. In other cases you must use LZMA_FINISH_ANY. - - If LZMA decoder sees end marker before reaching output limit, it returns SZ_OK, - and output value of destLen will be less than output buffer size limit. - You can check status result also. - - You can use multiple checks to test data integrity after full decompression: - 1) Check Result and "status" variable. - 2) Check that output(destLen) = uncompressedSize, if you know real uncompressedSize. - 3) Check that output(srcLen) = compressedSize, if you know real compressedSize. - You must use correct finish mode in that case. */ - -typedef enum -{ - LZMA_STATUS_NOT_SPECIFIED, /* use main error code instead */ - LZMA_STATUS_FINISHED_WITH_MARK, /* stream was finished with end mark. */ - LZMA_STATUS_NOT_FINISHED, /* stream was not finished */ - LZMA_STATUS_NEEDS_MORE_INPUT, /* you must provide more input bytes */ - LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK /* there is probability that stream was finished without end mark */ -} ELzmaStatus; - -/* ELzmaStatus is used only as output value for function call */ - - -/* ---------- Interfaces ---------- */ - -/* There are 3 levels of interfaces: - 1) Dictionary Interface - 2) Buffer Interface - 3) One Call Interface - You can select any of these interfaces, but don't mix functions from different - groups for same object. */ - - -/* There are two variants to allocate state for Dictionary Interface: - 1) LzmaDec_Allocate / LzmaDec_Free - 2) LzmaDec_AllocateProbs / LzmaDec_FreeProbs - You can use variant 2, if you set dictionary buffer manually. - For Buffer Interface you must always use variant 1. - -LzmaDec_Allocate* can return: - SZ_OK - SZ_ERROR_MEM - Memory allocation error - SZ_ERROR_UNSUPPORTED - Unsupported properties -*/ - -SRes LzmaDec_AllocateProbs(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc); -void LzmaDec_FreeProbs(CLzmaDec *p, ISzAllocPtr alloc); - -SRes LzmaDec_Allocate(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc); -void LzmaDec_Free(CLzmaDec *p, ISzAllocPtr alloc); - -/* ---------- Dictionary Interface ---------- */ - -/* You can use it, if you want to eliminate the overhead for data copying from - dictionary to some other external buffer. - You must work with CLzmaDec variables directly in this interface. - - STEPS: - LzmaDec_Construct() - LzmaDec_Allocate() - for (each new stream) - { - LzmaDec_Init() - while (it needs more decompression) - { - LzmaDec_DecodeToDic() - use data from CLzmaDec::dic and update CLzmaDec::dicPos - } - } - LzmaDec_Free() -*/ - -/* LzmaDec_DecodeToDic - - The decoding to internal dictionary buffer (CLzmaDec::dic). - You must manually update CLzmaDec::dicPos, if it reaches CLzmaDec::dicBufSize !!! - -finishMode: - It has meaning only if the decoding reaches output limit (dicLimit). - LZMA_FINISH_ANY - Decode just dicLimit bytes. - LZMA_FINISH_END - Stream must be finished after dicLimit. - -Returns: - SZ_OK - status: - LZMA_STATUS_FINISHED_WITH_MARK - LZMA_STATUS_NOT_FINISHED - LZMA_STATUS_NEEDS_MORE_INPUT - LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK - SZ_ERROR_DATA - Data error - SZ_ERROR_FAIL - Some unexpected error: internal error of code, memory corruption or hardware failure -*/ - -SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit, - const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); - - -/* ---------- Buffer Interface ---------- */ - -/* It's zlib-like interface. - See LzmaDec_DecodeToDic description for information about STEPS and return results, - but you must use LzmaDec_DecodeToBuf instead of LzmaDec_DecodeToDic and you don't need - to work with CLzmaDec variables manually. - -finishMode: - It has meaning only if the decoding reaches output limit (*destLen). - LZMA_FINISH_ANY - Decode just destLen bytes. - LZMA_FINISH_END - Stream must be finished after (*destLen). -*/ - -SRes LzmaDec_DecodeToBuf(CLzmaDec *p, Byte *dest, SizeT *destLen, - const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); - - -/* ---------- One Call Interface ---------- */ - -/* LzmaDecode - -finishMode: - It has meaning only if the decoding reaches output limit (*destLen). - LZMA_FINISH_ANY - Decode just destLen bytes. - LZMA_FINISH_END - Stream must be finished after (*destLen). - -Returns: - SZ_OK - status: - LZMA_STATUS_FINISHED_WITH_MARK - LZMA_STATUS_NOT_FINISHED - LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK - SZ_ERROR_DATA - Data error - SZ_ERROR_MEM - Memory allocation error - SZ_ERROR_UNSUPPORTED - Unsupported properties - SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src). - SZ_ERROR_FAIL - Some unexpected error: internal error of code, memory corruption or hardware failure -*/ - -SRes LzmaDecode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, - const Byte *propData, unsigned propSize, ELzmaFinishMode finishMode, - ELzmaStatus *status, ISzAllocPtr alloc); - -EXTERN_C_END - -#endif +/* LzmaDec.h -- LZMA Decoder +2020-03-19 : Igor Pavlov : Public domain */ + +#ifndef __LZMA_DEC_H +#define __LZMA_DEC_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +/* #define _LZMA_PROB32 */ +/* _LZMA_PROB32 can increase the speed on some CPUs, + but memory usage for CLzmaDec::probs will be doubled in that case */ + +typedef +#ifdef _LZMA_PROB32 + UInt32 +#else + UInt16 +#endif + CLzmaProb; + + +/* ---------- LZMA Properties ---------- */ + +#define LZMA_PROPS_SIZE 5 + +typedef struct _CLzmaProps +{ + Byte lc; + Byte lp; + Byte pb; + Byte _pad_; + UInt32 dicSize; +} CLzmaProps; + +/* LzmaProps_Decode - decodes properties +Returns: + SZ_OK + SZ_ERROR_UNSUPPORTED - Unsupported properties +*/ + +SRes LzmaProps_Decode(CLzmaProps *p, const Byte *data, unsigned size); + + +/* ---------- LZMA Decoder state ---------- */ + +/* LZMA_REQUIRED_INPUT_MAX = number of required input bytes for worst case. + Num bits = log2((2^11 / 31) ^ 22) + 26 < 134 + 26 = 160; */ + +#define LZMA_REQUIRED_INPUT_MAX 20 + +typedef struct +{ + /* Don't change this structure. ASM code can use it. */ + CLzmaProps prop; + CLzmaProb *probs; + CLzmaProb *probs_1664; + Byte *dic; + SizeT dicBufSize; + SizeT dicPos; + const Byte *buf; + UInt32 range; + UInt32 code; + UInt32 processedPos; + UInt32 checkDicSize; + UInt32 reps[4]; + UInt32 state; + UInt32 remainLen; + + UInt32 numProbs; + unsigned tempBufSize; + Byte tempBuf[LZMA_REQUIRED_INPUT_MAX]; +} CLzmaDec; + +#define LzmaDec_Construct(p) { (p)->dic = NULL; (p)->probs = NULL; } + +void LzmaDec_Init(CLzmaDec *p); + +/* There are two types of LZMA streams: + - Stream with end mark. That end mark adds about 6 bytes to compressed size. + - Stream without end mark. You must know exact uncompressed size to decompress such stream. */ + +typedef enum +{ + LZMA_FINISH_ANY, /* finish at any point */ + LZMA_FINISH_END /* block must be finished at the end */ +} ELzmaFinishMode; + +/* ELzmaFinishMode has meaning only if the decoding reaches output limit !!! + + You must use LZMA_FINISH_END, when you know that current output buffer + covers last bytes of block. In other cases you must use LZMA_FINISH_ANY. + + If LZMA decoder sees end marker before reaching output limit, it returns SZ_OK, + and output value of destLen will be less than output buffer size limit. + You can check status result also. + + You can use multiple checks to test data integrity after full decompression: + 1) Check Result and "status" variable. + 2) Check that output(destLen) = uncompressedSize, if you know real uncompressedSize. + 3) Check that output(srcLen) = compressedSize, if you know real compressedSize. + You must use correct finish mode in that case. */ + +typedef enum +{ + LZMA_STATUS_NOT_SPECIFIED, /* use main error code instead */ + LZMA_STATUS_FINISHED_WITH_MARK, /* stream was finished with end mark. */ + LZMA_STATUS_NOT_FINISHED, /* stream was not finished */ + LZMA_STATUS_NEEDS_MORE_INPUT, /* you must provide more input bytes */ + LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK /* there is probability that stream was finished without end mark */ +} ELzmaStatus; + +/* ELzmaStatus is used only as output value for function call */ + + +/* ---------- Interfaces ---------- */ + +/* There are 3 levels of interfaces: + 1) Dictionary Interface + 2) Buffer Interface + 3) One Call Interface + You can select any of these interfaces, but don't mix functions from different + groups for same object. */ + + +/* There are two variants to allocate state for Dictionary Interface: + 1) LzmaDec_Allocate / LzmaDec_Free + 2) LzmaDec_AllocateProbs / LzmaDec_FreeProbs + You can use variant 2, if you set dictionary buffer manually. + For Buffer Interface you must always use variant 1. + +LzmaDec_Allocate* can return: + SZ_OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_UNSUPPORTED - Unsupported properties +*/ + +SRes LzmaDec_AllocateProbs(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc); +void LzmaDec_FreeProbs(CLzmaDec *p, ISzAllocPtr alloc); + +SRes LzmaDec_Allocate(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc); +void LzmaDec_Free(CLzmaDec *p, ISzAllocPtr alloc); + +/* ---------- Dictionary Interface ---------- */ + +/* You can use it, if you want to eliminate the overhead for data copying from + dictionary to some other external buffer. + You must work with CLzmaDec variables directly in this interface. + + STEPS: + LzmaDec_Construct() + LzmaDec_Allocate() + for (each new stream) + { + LzmaDec_Init() + while (it needs more decompression) + { + LzmaDec_DecodeToDic() + use data from CLzmaDec::dic and update CLzmaDec::dicPos + } + } + LzmaDec_Free() +*/ + +/* LzmaDec_DecodeToDic + + The decoding to internal dictionary buffer (CLzmaDec::dic). + You must manually update CLzmaDec::dicPos, if it reaches CLzmaDec::dicBufSize !!! + +finishMode: + It has meaning only if the decoding reaches output limit (dicLimit). + LZMA_FINISH_ANY - Decode just dicLimit bytes. + LZMA_FINISH_END - Stream must be finished after dicLimit. + +Returns: + SZ_OK + status: + LZMA_STATUS_FINISHED_WITH_MARK + LZMA_STATUS_NOT_FINISHED + LZMA_STATUS_NEEDS_MORE_INPUT + LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK + SZ_ERROR_DATA - Data error + SZ_ERROR_FAIL - Some unexpected error: internal error of code, memory corruption or hardware failure +*/ + +SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit, + const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); + + +/* ---------- Buffer Interface ---------- */ + +/* It's zlib-like interface. + See LzmaDec_DecodeToDic description for information about STEPS and return results, + but you must use LzmaDec_DecodeToBuf instead of LzmaDec_DecodeToDic and you don't need + to work with CLzmaDec variables manually. + +finishMode: + It has meaning only if the decoding reaches output limit (*destLen). + LZMA_FINISH_ANY - Decode just destLen bytes. + LZMA_FINISH_END - Stream must be finished after (*destLen). +*/ + +SRes LzmaDec_DecodeToBuf(CLzmaDec *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); + + +/* ---------- One Call Interface ---------- */ + +/* LzmaDecode + +finishMode: + It has meaning only if the decoding reaches output limit (*destLen). + LZMA_FINISH_ANY - Decode just destLen bytes. + LZMA_FINISH_END - Stream must be finished after (*destLen). + +Returns: + SZ_OK + status: + LZMA_STATUS_FINISHED_WITH_MARK + LZMA_STATUS_NOT_FINISHED + LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK + SZ_ERROR_DATA - Data error + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_UNSUPPORTED - Unsupported properties + SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src). + SZ_ERROR_FAIL - Some unexpected error: internal error of code, memory corruption or hardware failure +*/ + +SRes LzmaDecode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + const Byte *propData, unsigned propSize, ELzmaFinishMode finishMode, + ELzmaStatus *status, ISzAllocPtr alloc); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/LzmaEnc.c b/deps/LZMA-SDK/C/LzmaEnc.c index ca9154aef..c8b31a19e 100644 --- a/deps/LZMA-SDK/C/LzmaEnc.c +++ b/deps/LZMA-SDK/C/LzmaEnc.c @@ -1,3162 +1,3162 @@ -/* LzmaEnc.c -- LZMA Encoder -2022-07-15: Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include - -/* #define SHOW_STAT */ -/* #define SHOW_STAT2 */ - -#if defined(SHOW_STAT) || defined(SHOW_STAT2) -#include -#endif - -#include "CpuArch.h" -#include "LzmaEnc.h" - -#include "LzFind.h" -#ifndef _7ZIP_ST -#include "LzFindMt.h" -#endif - -/* the following LzmaEnc_* declarations is internal LZMA interface for LZMA2 encoder */ - -SRes LzmaEnc_PrepareForLzma2(CLzmaEncHandle pp, ISeqInStream *inStream, UInt32 keepWindowSize, - ISzAllocPtr alloc, ISzAllocPtr allocBig); -SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen, - UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig); -SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit, - Byte *dest, size_t *destLen, UInt32 desiredPackSize, UInt32 *unpackSize); -const Byte *LzmaEnc_GetCurBuf(CLzmaEncHandle pp); -void LzmaEnc_Finish(CLzmaEncHandle pp); -void LzmaEnc_SaveState(CLzmaEncHandle pp); -void LzmaEnc_RestoreState(CLzmaEncHandle pp); - -#ifdef SHOW_STAT -static unsigned g_STAT_OFFSET = 0; -#endif - -/* for good normalization speed we still reserve 256 MB before 4 GB range */ -#define kLzmaMaxHistorySize ((UInt32)15 << 28) - -#define kNumTopBits 24 -#define kTopValue ((UInt32)1 << kNumTopBits) - -#define kNumBitModelTotalBits 11 -#define kBitModelTotal (1 << kNumBitModelTotalBits) -#define kNumMoveBits 5 -#define kProbInitValue (kBitModelTotal >> 1) - -#define kNumMoveReducingBits 4 -#define kNumBitPriceShiftBits 4 -// #define kBitPrice (1 << kNumBitPriceShiftBits) - -#define REP_LEN_COUNT 64 - -void LzmaEncProps_Init(CLzmaEncProps *p) -{ - p->level = 5; - p->dictSize = p->mc = 0; - p->reduceSize = (UInt64)(Int64)-1; - p->lc = p->lp = p->pb = p->algo = p->fb = p->btMode = p->numHashBytes = p->numThreads = -1; - p->writeEndMark = 0; - p->affinity = 0; -} - -void LzmaEncProps_Normalize(CLzmaEncProps *p) -{ - int level = p->level; - if (level < 0) level = 5; - p->level = level; - - if (p->dictSize == 0) - p->dictSize = - ( level <= 3 ? ((UInt32)1 << (level * 2 + 16)) : - ( level <= 6 ? ((UInt32)1 << (level + 19)) : - ( level <= 7 ? ((UInt32)1 << 25) : ((UInt32)1 << 26) - ))); - - if (p->dictSize > p->reduceSize) - { - UInt32 v = (UInt32)p->reduceSize; - const UInt32 kReduceMin = ((UInt32)1 << 12); - if (v < kReduceMin) - v = kReduceMin; - if (p->dictSize > v) - p->dictSize = v; - } - - if (p->lc < 0) p->lc = 3; - if (p->lp < 0) p->lp = 0; - if (p->pb < 0) p->pb = 2; - - if (p->algo < 0) p->algo = (level < 5 ? 0 : 1); - if (p->fb < 0) p->fb = (level < 7 ? 32 : 64); - if (p->btMode < 0) p->btMode = (p->algo == 0 ? 0 : 1); - if (p->numHashBytes < 0) p->numHashBytes = (p->btMode ? 4 : 5); - if (p->mc == 0) p->mc = (16 + ((unsigned)p->fb >> 1)) >> (p->btMode ? 0 : 1); - - if (p->numThreads < 0) - p->numThreads = - #ifndef _7ZIP_ST - ((p->btMode && p->algo) ? 2 : 1); - #else - 1; - #endif -} - -UInt32 LzmaEncProps_GetDictSize(const CLzmaEncProps *props2) -{ - CLzmaEncProps props = *props2; - LzmaEncProps_Normalize(&props); - return props.dictSize; -} - - -/* -x86/x64: - -BSR: - IF (SRC == 0) ZF = 1, DEST is undefined; - AMD : DEST is unchanged; - IF (SRC != 0) ZF = 0; DEST is index of top non-zero bit - BSR is slow in some processors - -LZCNT: - IF (SRC == 0) CF = 1, DEST is size_in_bits_of_register(src) (32 or 64) - IF (SRC != 0) CF = 0, DEST = num_lead_zero_bits - IF (DEST == 0) ZF = 1; - -LZCNT works only in new processors starting from Haswell. -if LZCNT is not supported by processor, then it's executed as BSR. -LZCNT can be faster than BSR, if supported. -*/ - -// #define LZMA_LOG_BSR - -#if defined(MY_CPU_ARM_OR_ARM64) /* || defined(MY_CPU_X86_OR_AMD64) */ - - #if (defined(__clang__) && (__clang_major__ >= 6)) \ - || (defined(__GNUC__) && (__GNUC__ >= 6)) - #define LZMA_LOG_BSR - #elif defined(_MSC_VER) && (_MSC_VER >= 1300) - // #if defined(MY_CPU_ARM_OR_ARM64) - #define LZMA_LOG_BSR - // #endif - #endif -#endif - -// #include - -#ifdef LZMA_LOG_BSR - -#if defined(__clang__) \ - || defined(__GNUC__) - -/* - C code: : (30 - __builtin_clz(x)) - gcc9/gcc10 for x64 /x86 : 30 - (bsr(x) xor 31) - clang10 for x64 : 31 + (bsr(x) xor -32) -*/ - - #define MY_clz(x) ((unsigned)__builtin_clz(x)) - // __lzcnt32 - // __builtin_ia32_lzcnt_u32 - -#else // #if defined(_MSC_VER) - - #ifdef MY_CPU_ARM_OR_ARM64 - - #define MY_clz _CountLeadingZeros - - #else // if defined(MY_CPU_X86_OR_AMD64) - - // #define MY_clz __lzcnt // we can use lzcnt (unsupported by old CPU) - // _BitScanReverse code is not optimal for some MSVC compilers - #define BSR2_RET(pos, res) { unsigned long zz; _BitScanReverse(&zz, (pos)); zz--; \ - res = (zz + zz) + (pos >> zz); } - - #endif // MY_CPU_X86_OR_AMD64 - -#endif // _MSC_VER - - -#ifndef BSR2_RET - - #define BSR2_RET(pos, res) { unsigned zz = 30 - MY_clz(pos); \ - res = (zz + zz) + (pos >> zz); } - -#endif - - -unsigned GetPosSlot1(UInt32 pos); -unsigned GetPosSlot1(UInt32 pos) -{ - unsigned res; - BSR2_RET(pos, res); - return res; -} -#define GetPosSlot2(pos, res) { BSR2_RET(pos, res); } -#define GetPosSlot(pos, res) { if (pos < 2) res = pos; else BSR2_RET(pos, res); } - - -#else // ! LZMA_LOG_BSR - -#define kNumLogBits (11 + sizeof(size_t) / 8 * 3) - -#define kDicLogSizeMaxCompress ((kNumLogBits - 1) * 2 + 7) - -static void LzmaEnc_FastPosInit(Byte *g_FastPos) -{ - unsigned slot; - g_FastPos[0] = 0; - g_FastPos[1] = 1; - g_FastPos += 2; - - for (slot = 2; slot < kNumLogBits * 2; slot++) - { - size_t k = ((size_t)1 << ((slot >> 1) - 1)); - size_t j; - for (j = 0; j < k; j++) - g_FastPos[j] = (Byte)slot; - g_FastPos += k; - } -} - -/* we can use ((limit - pos) >> 31) only if (pos < ((UInt32)1 << 31)) */ -/* -#define BSR2_RET(pos, res) { unsigned zz = 6 + ((kNumLogBits - 1) & \ - (0 - (((((UInt32)1 << (kNumLogBits + 6)) - 1) - pos) >> 31))); \ - res = p->g_FastPos[pos >> zz] + (zz * 2); } -*/ - -/* -#define BSR2_RET(pos, res) { unsigned zz = 6 + ((kNumLogBits - 1) & \ - (0 - (((((UInt32)1 << (kNumLogBits)) - 1) - (pos >> 6)) >> 31))); \ - res = p->g_FastPos[pos >> zz] + (zz * 2); } -*/ - -#define BSR2_RET(pos, res) { unsigned zz = (pos < (1 << (kNumLogBits + 6))) ? 6 : 6 + kNumLogBits - 1; \ - res = p->g_FastPos[pos >> zz] + (zz * 2); } - -/* -#define BSR2_RET(pos, res) { res = (pos < (1 << (kNumLogBits + 6))) ? \ - p->g_FastPos[pos >> 6] + 12 : \ - p->g_FastPos[pos >> (6 + kNumLogBits - 1)] + (6 + (kNumLogBits - 1)) * 2; } -*/ - -#define GetPosSlot1(pos) p->g_FastPos[pos] -#define GetPosSlot2(pos, res) { BSR2_RET(pos, res); } -#define GetPosSlot(pos, res) { if (pos < kNumFullDistances) res = p->g_FastPos[pos & (kNumFullDistances - 1)]; else BSR2_RET(pos, res); } - -#endif // LZMA_LOG_BSR - - -#define LZMA_NUM_REPS 4 - -typedef UInt16 CState; -typedef UInt16 CExtra; - -typedef struct -{ - UInt32 price; - CState state; - CExtra extra; - // 0 : normal - // 1 : LIT : MATCH - // > 1 : MATCH (extra-1) : LIT : REP0 (len) - UInt32 len; - UInt32 dist; - UInt32 reps[LZMA_NUM_REPS]; -} COptimal; - - -// 18.06 -#define kNumOpts (1 << 11) -#define kPackReserve (kNumOpts * 8) -// #define kNumOpts (1 << 12) -// #define kPackReserve (1 + kNumOpts * 2) - -#define kNumLenToPosStates 4 -#define kNumPosSlotBits 6 -// #define kDicLogSizeMin 0 -#define kDicLogSizeMax 32 -#define kDistTableSizeMax (kDicLogSizeMax * 2) - -#define kNumAlignBits 4 -#define kAlignTableSize (1 << kNumAlignBits) -#define kAlignMask (kAlignTableSize - 1) - -#define kStartPosModelIndex 4 -#define kEndPosModelIndex 14 -#define kNumFullDistances (1 << (kEndPosModelIndex >> 1)) - -typedef -#ifdef _LZMA_PROB32 - UInt32 -#else - UInt16 -#endif - CLzmaProb; - -#define LZMA_PB_MAX 4 -#define LZMA_LC_MAX 8 -#define LZMA_LP_MAX 4 - -#define LZMA_NUM_PB_STATES_MAX (1 << LZMA_PB_MAX) - -#define kLenNumLowBits 3 -#define kLenNumLowSymbols (1 << kLenNumLowBits) -#define kLenNumHighBits 8 -#define kLenNumHighSymbols (1 << kLenNumHighBits) -#define kLenNumSymbolsTotal (kLenNumLowSymbols * 2 + kLenNumHighSymbols) - -#define LZMA_MATCH_LEN_MIN 2 -#define LZMA_MATCH_LEN_MAX (LZMA_MATCH_LEN_MIN + kLenNumSymbolsTotal - 1) - -#define kNumStates 12 - - -typedef struct -{ - CLzmaProb low[LZMA_NUM_PB_STATES_MAX << (kLenNumLowBits + 1)]; - CLzmaProb high[kLenNumHighSymbols]; -} CLenEnc; - - -typedef struct -{ - unsigned tableSize; - UInt32 prices[LZMA_NUM_PB_STATES_MAX][kLenNumSymbolsTotal]; - // UInt32 prices1[LZMA_NUM_PB_STATES_MAX][kLenNumLowSymbols * 2]; - // UInt32 prices2[kLenNumSymbolsTotal]; -} CLenPriceEnc; - -#define GET_PRICE_LEN(p, posState, len) \ - ((p)->prices[posState][(size_t)(len) - LZMA_MATCH_LEN_MIN]) - -/* -#define GET_PRICE_LEN(p, posState, len) \ - ((p)->prices2[(size_t)(len) - 2] + ((p)->prices1[posState][((len) - 2) & (kLenNumLowSymbols * 2 - 1)] & (((len) - 2 - kLenNumLowSymbols * 2) >> 9))) -*/ - -typedef struct -{ - UInt32 range; - unsigned cache; - UInt64 low; - UInt64 cacheSize; - Byte *buf; - Byte *bufLim; - Byte *bufBase; - ISeqOutStream *outStream; - UInt64 processed; - SRes res; -} CRangeEnc; - - -typedef struct -{ - CLzmaProb *litProbs; - - unsigned state; - UInt32 reps[LZMA_NUM_REPS]; - - CLzmaProb posAlignEncoder[1 << kNumAlignBits]; - CLzmaProb isRep[kNumStates]; - CLzmaProb isRepG0[kNumStates]; - CLzmaProb isRepG1[kNumStates]; - CLzmaProb isRepG2[kNumStates]; - CLzmaProb isMatch[kNumStates][LZMA_NUM_PB_STATES_MAX]; - CLzmaProb isRep0Long[kNumStates][LZMA_NUM_PB_STATES_MAX]; - - CLzmaProb posSlotEncoder[kNumLenToPosStates][1 << kNumPosSlotBits]; - CLzmaProb posEncoders[kNumFullDistances]; - - CLenEnc lenProbs; - CLenEnc repLenProbs; - -} CSaveState; - - -typedef UInt32 CProbPrice; - - -typedef struct -{ - void *matchFinderObj; - IMatchFinder2 matchFinder; - - unsigned optCur; - unsigned optEnd; - - unsigned longestMatchLen; - unsigned numPairs; - UInt32 numAvail; - - unsigned state; - unsigned numFastBytes; - unsigned additionalOffset; - UInt32 reps[LZMA_NUM_REPS]; - unsigned lpMask, pbMask; - CLzmaProb *litProbs; - CRangeEnc rc; - - UInt32 backRes; - - unsigned lc, lp, pb; - unsigned lclp; - - BoolInt fastMode; - BoolInt writeEndMark; - BoolInt finished; - BoolInt multiThread; - BoolInt needInit; - // BoolInt _maxMode; - - UInt64 nowPos64; - - unsigned matchPriceCount; - // unsigned alignPriceCount; - int repLenEncCounter; - - unsigned distTableSize; - - UInt32 dictSize; - SRes result; - - #ifndef _7ZIP_ST - BoolInt mtMode; - // begin of CMatchFinderMt is used in LZ thread - CMatchFinderMt matchFinderMt; - // end of CMatchFinderMt is used in BT and HASH threads - // #else - // CMatchFinder matchFinderBase; - #endif - CMatchFinder matchFinderBase; - - - // we suppose that we have 8-bytes alignment after CMatchFinder - - #ifndef _7ZIP_ST - Byte pad[128]; - #endif - - // LZ thread - CProbPrice ProbPrices[kBitModelTotal >> kNumMoveReducingBits]; - - // we want {len , dist} pairs to be 8-bytes aligned in matches array - UInt32 matches[LZMA_MATCH_LEN_MAX * 2 + 2]; - - // we want 8-bytes alignment here - UInt32 alignPrices[kAlignTableSize]; - UInt32 posSlotPrices[kNumLenToPosStates][kDistTableSizeMax]; - UInt32 distancesPrices[kNumLenToPosStates][kNumFullDistances]; - - CLzmaProb posAlignEncoder[1 << kNumAlignBits]; - CLzmaProb isRep[kNumStates]; - CLzmaProb isRepG0[kNumStates]; - CLzmaProb isRepG1[kNumStates]; - CLzmaProb isRepG2[kNumStates]; - CLzmaProb isMatch[kNumStates][LZMA_NUM_PB_STATES_MAX]; - CLzmaProb isRep0Long[kNumStates][LZMA_NUM_PB_STATES_MAX]; - CLzmaProb posSlotEncoder[kNumLenToPosStates][1 << kNumPosSlotBits]; - CLzmaProb posEncoders[kNumFullDistances]; - - CLenEnc lenProbs; - CLenEnc repLenProbs; - - #ifndef LZMA_LOG_BSR - Byte g_FastPos[1 << kNumLogBits]; - #endif - - CLenPriceEnc lenEnc; - CLenPriceEnc repLenEnc; - - COptimal opt[kNumOpts]; - - CSaveState saveState; - - // BoolInt mf_Failure; - #ifndef _7ZIP_ST - Byte pad2[128]; - #endif -} CLzmaEnc; - - -#define MFB (p->matchFinderBase) -/* -#ifndef _7ZIP_ST -#define MFB (p->matchFinderMt.MatchFinder) -#endif -*/ - -#define COPY_ARR(dest, src, arr) memcpy(dest->arr, src->arr, sizeof(src->arr)); - -void LzmaEnc_SaveState(CLzmaEncHandle pp) -{ - CLzmaEnc *p = (CLzmaEnc *)pp; - CSaveState *dest = &p->saveState; - - dest->state = p->state; - - dest->lenProbs = p->lenProbs; - dest->repLenProbs = p->repLenProbs; - - COPY_ARR(dest, p, reps); - - COPY_ARR(dest, p, posAlignEncoder); - COPY_ARR(dest, p, isRep); - COPY_ARR(dest, p, isRepG0); - COPY_ARR(dest, p, isRepG1); - COPY_ARR(dest, p, isRepG2); - COPY_ARR(dest, p, isMatch); - COPY_ARR(dest, p, isRep0Long); - COPY_ARR(dest, p, posSlotEncoder); - COPY_ARR(dest, p, posEncoders); - - memcpy(dest->litProbs, p->litProbs, ((UInt32)0x300 << p->lclp) * sizeof(CLzmaProb)); -} - - -void LzmaEnc_RestoreState(CLzmaEncHandle pp) -{ - CLzmaEnc *dest = (CLzmaEnc *)pp; - const CSaveState *p = &dest->saveState; - - dest->state = p->state; - - dest->lenProbs = p->lenProbs; - dest->repLenProbs = p->repLenProbs; - - COPY_ARR(dest, p, reps); - - COPY_ARR(dest, p, posAlignEncoder); - COPY_ARR(dest, p, isRep); - COPY_ARR(dest, p, isRepG0); - COPY_ARR(dest, p, isRepG1); - COPY_ARR(dest, p, isRepG2); - COPY_ARR(dest, p, isMatch); - COPY_ARR(dest, p, isRep0Long); - COPY_ARR(dest, p, posSlotEncoder); - COPY_ARR(dest, p, posEncoders); - - memcpy(dest->litProbs, p->litProbs, ((UInt32)0x300 << dest->lclp) * sizeof(CLzmaProb)); -} - - - -SRes LzmaEnc_SetProps(CLzmaEncHandle pp, const CLzmaEncProps *props2) -{ - CLzmaEnc *p = (CLzmaEnc *)pp; - CLzmaEncProps props = *props2; - LzmaEncProps_Normalize(&props); - - if (props.lc > LZMA_LC_MAX - || props.lp > LZMA_LP_MAX - || props.pb > LZMA_PB_MAX) - return SZ_ERROR_PARAM; - - - if (props.dictSize > kLzmaMaxHistorySize) - props.dictSize = kLzmaMaxHistorySize; - - #ifndef LZMA_LOG_BSR - { - const UInt64 dict64 = props.dictSize; - if (dict64 > ((UInt64)1 << kDicLogSizeMaxCompress)) - return SZ_ERROR_PARAM; - } - #endif - - p->dictSize = props.dictSize; - { - unsigned fb = (unsigned)props.fb; - if (fb < 5) - fb = 5; - if (fb > LZMA_MATCH_LEN_MAX) - fb = LZMA_MATCH_LEN_MAX; - p->numFastBytes = fb; - } - p->lc = (unsigned)props.lc; - p->lp = (unsigned)props.lp; - p->pb = (unsigned)props.pb; - p->fastMode = (props.algo == 0); - // p->_maxMode = True; - MFB.btMode = (Byte)(props.btMode ? 1 : 0); - { - unsigned numHashBytes = 4; - if (props.btMode) - { - if (props.numHashBytes < 2) numHashBytes = 2; - else if (props.numHashBytes < 4) numHashBytes = (unsigned)props.numHashBytes; - } - if (props.numHashBytes >= 5) numHashBytes = 5; - - MFB.numHashBytes = numHashBytes; - } - - MFB.cutValue = props.mc; - - p->writeEndMark = (BoolInt)props.writeEndMark; - - #ifndef _7ZIP_ST - /* - if (newMultiThread != _multiThread) - { - ReleaseMatchFinder(); - _multiThread = newMultiThread; - } - */ - p->multiThread = (props.numThreads > 1); - p->matchFinderMt.btSync.affinity = - p->matchFinderMt.hashSync.affinity = props.affinity; - #endif - - return SZ_OK; -} - - -void LzmaEnc_SetDataSize(CLzmaEncHandle pp, UInt64 expectedDataSiize) -{ - CLzmaEnc *p = (CLzmaEnc *)pp; - MFB.expectedDataSize = expectedDataSiize; -} - - -#define kState_Start 0 -#define kState_LitAfterMatch 4 -#define kState_LitAfterRep 5 -#define kState_MatchAfterLit 7 -#define kState_RepAfterLit 8 - -static const Byte kLiteralNextStates[kNumStates] = {0, 0, 0, 0, 1, 2, 3, 4, 5, 6, 4, 5}; -static const Byte kMatchNextStates[kNumStates] = {7, 7, 7, 7, 7, 7, 7, 10, 10, 10, 10, 10}; -static const Byte kRepNextStates[kNumStates] = {8, 8, 8, 8, 8, 8, 8, 11, 11, 11, 11, 11}; -static const Byte kShortRepNextStates[kNumStates]= {9, 9, 9, 9, 9, 9, 9, 11, 11, 11, 11, 11}; - -#define IsLitState(s) ((s) < 7) -#define GetLenToPosState2(len) (((len) < kNumLenToPosStates - 1) ? (len) : kNumLenToPosStates - 1) -#define GetLenToPosState(len) (((len) < kNumLenToPosStates + 1) ? (len) - 2 : kNumLenToPosStates - 1) - -#define kInfinityPrice (1 << 30) - -static void RangeEnc_Construct(CRangeEnc *p) -{ - p->outStream = NULL; - p->bufBase = NULL; -} - -#define RangeEnc_GetProcessed(p) ( (p)->processed + (size_t)((p)->buf - (p)->bufBase) + (p)->cacheSize) -#define RangeEnc_GetProcessed_sizet(p) ((size_t)(p)->processed + (size_t)((p)->buf - (p)->bufBase) + (size_t)(p)->cacheSize) - -#define RC_BUF_SIZE (1 << 16) - -static int RangeEnc_Alloc(CRangeEnc *p, ISzAllocPtr alloc) -{ - if (!p->bufBase) - { - p->bufBase = (Byte *)ISzAlloc_Alloc(alloc, RC_BUF_SIZE); - if (!p->bufBase) - return 0; - p->bufLim = p->bufBase + RC_BUF_SIZE; - } - return 1; -} - -static void RangeEnc_Free(CRangeEnc *p, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, p->bufBase); - p->bufBase = NULL; -} - -static void RangeEnc_Init(CRangeEnc *p) -{ - p->range = 0xFFFFFFFF; - p->cache = 0; - p->low = 0; - p->cacheSize = 0; - - p->buf = p->bufBase; - - p->processed = 0; - p->res = SZ_OK; -} - -MY_NO_INLINE static void RangeEnc_FlushStream(CRangeEnc *p) -{ - const size_t num = (size_t)(p->buf - p->bufBase); - if (p->res == SZ_OK) - { - if (num != ISeqOutStream_Write(p->outStream, p->bufBase, num)) - p->res = SZ_ERROR_WRITE; - } - p->processed += num; - p->buf = p->bufBase; -} - -MY_NO_INLINE static void MY_FAST_CALL RangeEnc_ShiftLow(CRangeEnc *p) -{ - UInt32 low = (UInt32)p->low; - unsigned high = (unsigned)(p->low >> 32); - p->low = (UInt32)(low << 8); - if (low < (UInt32)0xFF000000 || high != 0) - { - { - Byte *buf = p->buf; - *buf++ = (Byte)(p->cache + high); - p->cache = (unsigned)(low >> 24); - p->buf = buf; - if (buf == p->bufLim) - RangeEnc_FlushStream(p); - if (p->cacheSize == 0) - return; - } - high += 0xFF; - for (;;) - { - Byte *buf = p->buf; - *buf++ = (Byte)(high); - p->buf = buf; - if (buf == p->bufLim) - RangeEnc_FlushStream(p); - if (--p->cacheSize == 0) - return; - } - } - p->cacheSize++; -} - -static void RangeEnc_FlushData(CRangeEnc *p) -{ - int i; - for (i = 0; i < 5; i++) - RangeEnc_ShiftLow(p); -} - -#define RC_NORM(p) if (range < kTopValue) { range <<= 8; RangeEnc_ShiftLow(p); } - -#define RC_BIT_PRE(p, prob) \ - ttt = *(prob); \ - newBound = (range >> kNumBitModelTotalBits) * ttt; - -// #define _LZMA_ENC_USE_BRANCH - -#ifdef _LZMA_ENC_USE_BRANCH - -#define RC_BIT(p, prob, bit) { \ - RC_BIT_PRE(p, prob) \ - if (bit == 0) { range = newBound; ttt += (kBitModelTotal - ttt) >> kNumMoveBits; } \ - else { (p)->low += newBound; range -= newBound; ttt -= ttt >> kNumMoveBits; } \ - *(prob) = (CLzmaProb)ttt; \ - RC_NORM(p) \ - } - -#else - -#define RC_BIT(p, prob, bit) { \ - UInt32 mask; \ - RC_BIT_PRE(p, prob) \ - mask = 0 - (UInt32)bit; \ - range &= mask; \ - mask &= newBound; \ - range -= mask; \ - (p)->low += mask; \ - mask = (UInt32)bit - 1; \ - range += newBound & mask; \ - mask &= (kBitModelTotal - ((1 << kNumMoveBits) - 1)); \ - mask += ((1 << kNumMoveBits) - 1); \ - ttt += (UInt32)((Int32)(mask - ttt) >> kNumMoveBits); \ - *(prob) = (CLzmaProb)ttt; \ - RC_NORM(p) \ - } - -#endif - - - - -#define RC_BIT_0_BASE(p, prob) \ - range = newBound; *(prob) = (CLzmaProb)(ttt + ((kBitModelTotal - ttt) >> kNumMoveBits)); - -#define RC_BIT_1_BASE(p, prob) \ - range -= newBound; (p)->low += newBound; *(prob) = (CLzmaProb)(ttt - (ttt >> kNumMoveBits)); \ - -#define RC_BIT_0(p, prob) \ - RC_BIT_0_BASE(p, prob) \ - RC_NORM(p) - -#define RC_BIT_1(p, prob) \ - RC_BIT_1_BASE(p, prob) \ - RC_NORM(p) - -static void RangeEnc_EncodeBit_0(CRangeEnc *p, CLzmaProb *prob) -{ - UInt32 range, ttt, newBound; - range = p->range; - RC_BIT_PRE(p, prob) - RC_BIT_0(p, prob) - p->range = range; -} - -static void LitEnc_Encode(CRangeEnc *p, CLzmaProb *probs, UInt32 sym) -{ - UInt32 range = p->range; - sym |= 0x100; - do - { - UInt32 ttt, newBound; - // RangeEnc_EncodeBit(p, probs + (sym >> 8), (sym >> 7) & 1); - CLzmaProb *prob = probs + (sym >> 8); - UInt32 bit = (sym >> 7) & 1; - sym <<= 1; - RC_BIT(p, prob, bit); - } - while (sym < 0x10000); - p->range = range; -} - -static void LitEnc_EncodeMatched(CRangeEnc *p, CLzmaProb *probs, UInt32 sym, UInt32 matchByte) -{ - UInt32 range = p->range; - UInt32 offs = 0x100; - sym |= 0x100; - do - { - UInt32 ttt, newBound; - CLzmaProb *prob; - UInt32 bit; - matchByte <<= 1; - // RangeEnc_EncodeBit(p, probs + (offs + (matchByte & offs) + (sym >> 8)), (sym >> 7) & 1); - prob = probs + (offs + (matchByte & offs) + (sym >> 8)); - bit = (sym >> 7) & 1; - sym <<= 1; - offs &= ~(matchByte ^ sym); - RC_BIT(p, prob, bit); - } - while (sym < 0x10000); - p->range = range; -} - - - -static void LzmaEnc_InitPriceTables(CProbPrice *ProbPrices) -{ - UInt32 i; - for (i = 0; i < (kBitModelTotal >> kNumMoveReducingBits); i++) - { - const unsigned kCyclesBits = kNumBitPriceShiftBits; - UInt32 w = (i << kNumMoveReducingBits) + (1 << (kNumMoveReducingBits - 1)); - unsigned bitCount = 0; - unsigned j; - for (j = 0; j < kCyclesBits; j++) - { - w = w * w; - bitCount <<= 1; - while (w >= ((UInt32)1 << 16)) - { - w >>= 1; - bitCount++; - } - } - ProbPrices[i] = (CProbPrice)(((unsigned)kNumBitModelTotalBits << kCyclesBits) - 15 - bitCount); - // printf("\n%3d: %5d", i, ProbPrices[i]); - } -} - - -#define GET_PRICE(prob, bit) \ - p->ProbPrices[((prob) ^ (unsigned)(((-(int)(bit))) & (kBitModelTotal - 1))) >> kNumMoveReducingBits]; - -#define GET_PRICEa(prob, bit) \ - ProbPrices[((prob) ^ (unsigned)((-((int)(bit))) & (kBitModelTotal - 1))) >> kNumMoveReducingBits]; - -#define GET_PRICE_0(prob) p->ProbPrices[(prob) >> kNumMoveReducingBits] -#define GET_PRICE_1(prob) p->ProbPrices[((prob) ^ (kBitModelTotal - 1)) >> kNumMoveReducingBits] - -#define GET_PRICEa_0(prob) ProbPrices[(prob) >> kNumMoveReducingBits] -#define GET_PRICEa_1(prob) ProbPrices[((prob) ^ (kBitModelTotal - 1)) >> kNumMoveReducingBits] - - -static UInt32 LitEnc_GetPrice(const CLzmaProb *probs, UInt32 sym, const CProbPrice *ProbPrices) -{ - UInt32 price = 0; - sym |= 0x100; - do - { - unsigned bit = sym & 1; - sym >>= 1; - price += GET_PRICEa(probs[sym], bit); - } - while (sym >= 2); - return price; -} - - -static UInt32 LitEnc_Matched_GetPrice(const CLzmaProb *probs, UInt32 sym, UInt32 matchByte, const CProbPrice *ProbPrices) -{ - UInt32 price = 0; - UInt32 offs = 0x100; - sym |= 0x100; - do - { - matchByte <<= 1; - price += GET_PRICEa(probs[offs + (matchByte & offs) + (sym >> 8)], (sym >> 7) & 1); - sym <<= 1; - offs &= ~(matchByte ^ sym); - } - while (sym < 0x10000); - return price; -} - - -static void RcTree_ReverseEncode(CRangeEnc *rc, CLzmaProb *probs, unsigned numBits, unsigned sym) -{ - UInt32 range = rc->range; - unsigned m = 1; - do - { - UInt32 ttt, newBound; - unsigned bit = sym & 1; - // RangeEnc_EncodeBit(rc, probs + m, bit); - sym >>= 1; - RC_BIT(rc, probs + m, bit); - m = (m << 1) | bit; - } - while (--numBits); - rc->range = range; -} - - - -static void LenEnc_Init(CLenEnc *p) -{ - unsigned i; - for (i = 0; i < (LZMA_NUM_PB_STATES_MAX << (kLenNumLowBits + 1)); i++) - p->low[i] = kProbInitValue; - for (i = 0; i < kLenNumHighSymbols; i++) - p->high[i] = kProbInitValue; -} - -static void LenEnc_Encode(CLenEnc *p, CRangeEnc *rc, unsigned sym, unsigned posState) -{ - UInt32 range, ttt, newBound; - CLzmaProb *probs = p->low; - range = rc->range; - RC_BIT_PRE(rc, probs); - if (sym >= kLenNumLowSymbols) - { - RC_BIT_1(rc, probs); - probs += kLenNumLowSymbols; - RC_BIT_PRE(rc, probs); - if (sym >= kLenNumLowSymbols * 2) - { - RC_BIT_1(rc, probs); - rc->range = range; - // RcTree_Encode(rc, p->high, kLenNumHighBits, sym - kLenNumLowSymbols * 2); - LitEnc_Encode(rc, p->high, sym - kLenNumLowSymbols * 2); - return; - } - sym -= kLenNumLowSymbols; - } - - // RcTree_Encode(rc, probs + (posState << kLenNumLowBits), kLenNumLowBits, sym); - { - unsigned m; - unsigned bit; - RC_BIT_0(rc, probs); - probs += (posState << (1 + kLenNumLowBits)); - bit = (sym >> 2) ; RC_BIT(rc, probs + 1, bit); m = (1 << 1) + bit; - bit = (sym >> 1) & 1; RC_BIT(rc, probs + m, bit); m = (m << 1) + bit; - bit = sym & 1; RC_BIT(rc, probs + m, bit); - rc->range = range; - } -} - -static void SetPrices_3(const CLzmaProb *probs, UInt32 startPrice, UInt32 *prices, const CProbPrice *ProbPrices) -{ - unsigned i; - for (i = 0; i < 8; i += 2) - { - UInt32 price = startPrice; - UInt32 prob; - price += GET_PRICEa(probs[1 ], (i >> 2)); - price += GET_PRICEa(probs[2 + (i >> 2)], (i >> 1) & 1); - prob = probs[4 + (i >> 1)]; - prices[i ] = price + GET_PRICEa_0(prob); - prices[i + 1] = price + GET_PRICEa_1(prob); - } -} - - -MY_NO_INLINE static void MY_FAST_CALL LenPriceEnc_UpdateTables( - CLenPriceEnc *p, - unsigned numPosStates, - const CLenEnc *enc, - const CProbPrice *ProbPrices) -{ - UInt32 b; - - { - unsigned prob = enc->low[0]; - UInt32 a, c; - unsigned posState; - b = GET_PRICEa_1(prob); - a = GET_PRICEa_0(prob); - c = b + GET_PRICEa_0(enc->low[kLenNumLowSymbols]); - for (posState = 0; posState < numPosStates; posState++) - { - UInt32 *prices = p->prices[posState]; - const CLzmaProb *probs = enc->low + (posState << (1 + kLenNumLowBits)); - SetPrices_3(probs, a, prices, ProbPrices); - SetPrices_3(probs + kLenNumLowSymbols, c, prices + kLenNumLowSymbols, ProbPrices); - } - } - - /* - { - unsigned i; - UInt32 b; - a = GET_PRICEa_0(enc->low[0]); - for (i = 0; i < kLenNumLowSymbols; i++) - p->prices2[i] = a; - a = GET_PRICEa_1(enc->low[0]); - b = a + GET_PRICEa_0(enc->low[kLenNumLowSymbols]); - for (i = kLenNumLowSymbols; i < kLenNumLowSymbols * 2; i++) - p->prices2[i] = b; - a += GET_PRICEa_1(enc->low[kLenNumLowSymbols]); - } - */ - - // p->counter = numSymbols; - // p->counter = 64; - - { - unsigned i = p->tableSize; - - if (i > kLenNumLowSymbols * 2) - { - const CLzmaProb *probs = enc->high; - UInt32 *prices = p->prices[0] + kLenNumLowSymbols * 2; - i -= kLenNumLowSymbols * 2 - 1; - i >>= 1; - b += GET_PRICEa_1(enc->low[kLenNumLowSymbols]); - do - { - /* - p->prices2[i] = a + - // RcTree_GetPrice(enc->high, kLenNumHighBits, i - kLenNumLowSymbols * 2, ProbPrices); - LitEnc_GetPrice(probs, i - kLenNumLowSymbols * 2, ProbPrices); - */ - // UInt32 price = a + RcTree_GetPrice(probs, kLenNumHighBits - 1, sym, ProbPrices); - unsigned sym = --i + (1 << (kLenNumHighBits - 1)); - UInt32 price = b; - do - { - unsigned bit = sym & 1; - sym >>= 1; - price += GET_PRICEa(probs[sym], bit); - } - while (sym >= 2); - - { - unsigned prob = probs[(size_t)i + (1 << (kLenNumHighBits - 1))]; - prices[(size_t)i * 2 ] = price + GET_PRICEa_0(prob); - prices[(size_t)i * 2 + 1] = price + GET_PRICEa_1(prob); - } - } - while (i); - - { - unsigned posState; - size_t num = (p->tableSize - kLenNumLowSymbols * 2) * sizeof(p->prices[0][0]); - for (posState = 1; posState < numPosStates; posState++) - memcpy(p->prices[posState] + kLenNumLowSymbols * 2, p->prices[0] + kLenNumLowSymbols * 2, num); - } - } - } -} - -/* - #ifdef SHOW_STAT - g_STAT_OFFSET += num; - printf("\n MovePos %u", num); - #endif -*/ - -#define MOVE_POS(p, num) { \ - p->additionalOffset += (num); \ - p->matchFinder.Skip(p->matchFinderObj, (UInt32)(num)); } - - -static unsigned ReadMatchDistances(CLzmaEnc *p, unsigned *numPairsRes) -{ - unsigned numPairs; - - p->additionalOffset++; - p->numAvail = p->matchFinder.GetNumAvailableBytes(p->matchFinderObj); - { - const UInt32 *d = p->matchFinder.GetMatches(p->matchFinderObj, p->matches); - // if (!d) { p->mf_Failure = True; *numPairsRes = 0; return 0; } - numPairs = (unsigned)(d - p->matches); - } - *numPairsRes = numPairs; - - #ifdef SHOW_STAT - printf("\n i = %u numPairs = %u ", g_STAT_OFFSET, numPairs / 2); - g_STAT_OFFSET++; - { - unsigned i; - for (i = 0; i < numPairs; i += 2) - printf("%2u %6u | ", p->matches[i], p->matches[i + 1]); - } - #endif - - if (numPairs == 0) - return 0; - { - const unsigned len = p->matches[(size_t)numPairs - 2]; - if (len != p->numFastBytes) - return len; - { - UInt32 numAvail = p->numAvail; - if (numAvail > LZMA_MATCH_LEN_MAX) - numAvail = LZMA_MATCH_LEN_MAX; - { - const Byte *p1 = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; - const Byte *p2 = p1 + len; - const ptrdiff_t dif = (ptrdiff_t)-1 - (ptrdiff_t)p->matches[(size_t)numPairs - 1]; - const Byte *lim = p1 + numAvail; - for (; p2 != lim && *p2 == p2[dif]; p2++) - {} - return (unsigned)(p2 - p1); - } - } - } -} - -#define MARK_LIT ((UInt32)(Int32)-1) - -#define MakeAs_Lit(p) { (p)->dist = MARK_LIT; (p)->extra = 0; } -#define MakeAs_ShortRep(p) { (p)->dist = 0; (p)->extra = 0; } -#define IsShortRep(p) ((p)->dist == 0) - - -#define GetPrice_ShortRep(p, state, posState) \ - ( GET_PRICE_0(p->isRepG0[state]) + GET_PRICE_0(p->isRep0Long[state][posState])) - -#define GetPrice_Rep_0(p, state, posState) ( \ - GET_PRICE_1(p->isMatch[state][posState]) \ - + GET_PRICE_1(p->isRep0Long[state][posState])) \ - + GET_PRICE_1(p->isRep[state]) \ - + GET_PRICE_0(p->isRepG0[state]) - -MY_FORCE_INLINE -static UInt32 GetPrice_PureRep(const CLzmaEnc *p, unsigned repIndex, size_t state, size_t posState) -{ - UInt32 price; - UInt32 prob = p->isRepG0[state]; - if (repIndex == 0) - { - price = GET_PRICE_0(prob); - price += GET_PRICE_1(p->isRep0Long[state][posState]); - } - else - { - price = GET_PRICE_1(prob); - prob = p->isRepG1[state]; - if (repIndex == 1) - price += GET_PRICE_0(prob); - else - { - price += GET_PRICE_1(prob); - price += GET_PRICE(p->isRepG2[state], repIndex - 2); - } - } - return price; -} - - -static unsigned Backward(CLzmaEnc *p, unsigned cur) -{ - unsigned wr = cur + 1; - p->optEnd = wr; - - for (;;) - { - UInt32 dist = p->opt[cur].dist; - unsigned len = (unsigned)p->opt[cur].len; - unsigned extra = (unsigned)p->opt[cur].extra; - cur -= len; - - if (extra) - { - wr--; - p->opt[wr].len = (UInt32)len; - cur -= extra; - len = extra; - if (extra == 1) - { - p->opt[wr].dist = dist; - dist = MARK_LIT; - } - else - { - p->opt[wr].dist = 0; - len--; - wr--; - p->opt[wr].dist = MARK_LIT; - p->opt[wr].len = 1; - } - } - - if (cur == 0) - { - p->backRes = dist; - p->optCur = wr; - return len; - } - - wr--; - p->opt[wr].dist = dist; - p->opt[wr].len = (UInt32)len; - } -} - - - -#define LIT_PROBS(pos, prevByte) \ - (p->litProbs + (UInt32)3 * (((((pos) << 8) + (prevByte)) & p->lpMask) << p->lc)) - - -static unsigned GetOptimum(CLzmaEnc *p, UInt32 position) -{ - unsigned last, cur; - UInt32 reps[LZMA_NUM_REPS]; - unsigned repLens[LZMA_NUM_REPS]; - UInt32 *matches; - - { - UInt32 numAvail; - unsigned numPairs, mainLen, repMaxIndex, i, posState; - UInt32 matchPrice, repMatchPrice; - const Byte *data; - Byte curByte, matchByte; - - p->optCur = p->optEnd = 0; - - if (p->additionalOffset == 0) - mainLen = ReadMatchDistances(p, &numPairs); - else - { - mainLen = p->longestMatchLen; - numPairs = p->numPairs; - } - - numAvail = p->numAvail; - if (numAvail < 2) - { - p->backRes = MARK_LIT; - return 1; - } - if (numAvail > LZMA_MATCH_LEN_MAX) - numAvail = LZMA_MATCH_LEN_MAX; - - data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; - repMaxIndex = 0; - - for (i = 0; i < LZMA_NUM_REPS; i++) - { - unsigned len; - const Byte *data2; - reps[i] = p->reps[i]; - data2 = data - reps[i]; - if (data[0] != data2[0] || data[1] != data2[1]) - { - repLens[i] = 0; - continue; - } - for (len = 2; len < numAvail && data[len] == data2[len]; len++) - {} - repLens[i] = len; - if (len > repLens[repMaxIndex]) - repMaxIndex = i; - if (len == LZMA_MATCH_LEN_MAX) // 21.03 : optimization - break; - } - - if (repLens[repMaxIndex] >= p->numFastBytes) - { - unsigned len; - p->backRes = (UInt32)repMaxIndex; - len = repLens[repMaxIndex]; - MOVE_POS(p, len - 1) - return len; - } - - matches = p->matches; - #define MATCHES matches - // #define MATCHES p->matches - - if (mainLen >= p->numFastBytes) - { - p->backRes = MATCHES[(size_t)numPairs - 1] + LZMA_NUM_REPS; - MOVE_POS(p, mainLen - 1) - return mainLen; - } - - curByte = *data; - matchByte = *(data - reps[0]); - - last = repLens[repMaxIndex]; - if (last <= mainLen) - last = mainLen; - - if (last < 2 && curByte != matchByte) - { - p->backRes = MARK_LIT; - return 1; - } - - p->opt[0].state = (CState)p->state; - - posState = (position & p->pbMask); - - { - const CLzmaProb *probs = LIT_PROBS(position, *(data - 1)); - p->opt[1].price = GET_PRICE_0(p->isMatch[p->state][posState]) + - (!IsLitState(p->state) ? - LitEnc_Matched_GetPrice(probs, curByte, matchByte, p->ProbPrices) : - LitEnc_GetPrice(probs, curByte, p->ProbPrices)); - } - - MakeAs_Lit(&p->opt[1]); - - matchPrice = GET_PRICE_1(p->isMatch[p->state][posState]); - repMatchPrice = matchPrice + GET_PRICE_1(p->isRep[p->state]); - - // 18.06 - if (matchByte == curByte && repLens[0] == 0) - { - UInt32 shortRepPrice = repMatchPrice + GetPrice_ShortRep(p, p->state, posState); - if (shortRepPrice < p->opt[1].price) - { - p->opt[1].price = shortRepPrice; - MakeAs_ShortRep(&p->opt[1]); - } - if (last < 2) - { - p->backRes = p->opt[1].dist; - return 1; - } - } - - p->opt[1].len = 1; - - p->opt[0].reps[0] = reps[0]; - p->opt[0].reps[1] = reps[1]; - p->opt[0].reps[2] = reps[2]; - p->opt[0].reps[3] = reps[3]; - - // ---------- REP ---------- - - for (i = 0; i < LZMA_NUM_REPS; i++) - { - unsigned repLen = repLens[i]; - UInt32 price; - if (repLen < 2) - continue; - price = repMatchPrice + GetPrice_PureRep(p, i, p->state, posState); - do - { - UInt32 price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState, repLen); - COptimal *opt = &p->opt[repLen]; - if (price2 < opt->price) - { - opt->price = price2; - opt->len = (UInt32)repLen; - opt->dist = (UInt32)i; - opt->extra = 0; - } - } - while (--repLen >= 2); - } - - - // ---------- MATCH ---------- - { - unsigned len = repLens[0] + 1; - if (len <= mainLen) - { - unsigned offs = 0; - UInt32 normalMatchPrice = matchPrice + GET_PRICE_0(p->isRep[p->state]); - - if (len < 2) - len = 2; - else - while (len > MATCHES[offs]) - offs += 2; - - for (; ; len++) - { - COptimal *opt; - UInt32 dist = MATCHES[(size_t)offs + 1]; - UInt32 price = normalMatchPrice + GET_PRICE_LEN(&p->lenEnc, posState, len); - unsigned lenToPosState = GetLenToPosState(len); - - if (dist < kNumFullDistances) - price += p->distancesPrices[lenToPosState][dist & (kNumFullDistances - 1)]; - else - { - unsigned slot; - GetPosSlot2(dist, slot); - price += p->alignPrices[dist & kAlignMask]; - price += p->posSlotPrices[lenToPosState][slot]; - } - - opt = &p->opt[len]; - - if (price < opt->price) - { - opt->price = price; - opt->len = (UInt32)len; - opt->dist = dist + LZMA_NUM_REPS; - opt->extra = 0; - } - - if (len == MATCHES[offs]) - { - offs += 2; - if (offs == numPairs) - break; - } - } - } - } - - - cur = 0; - - #ifdef SHOW_STAT2 - /* if (position >= 0) */ - { - unsigned i; - printf("\n pos = %4X", position); - for (i = cur; i <= last; i++) - printf("\nprice[%4X] = %u", position - cur + i, p->opt[i].price); - } - #endif - } - - - - // ---------- Optimal Parsing ---------- - - for (;;) - { - unsigned numAvail; - UInt32 numAvailFull; - unsigned newLen, numPairs, prev, state, posState, startLen; - UInt32 litPrice, matchPrice, repMatchPrice; - BoolInt nextIsLit; - Byte curByte, matchByte; - const Byte *data; - COptimal *curOpt, *nextOpt; - - if (++cur == last) - break; - - // 18.06 - if (cur >= kNumOpts - 64) - { - unsigned j, best; - UInt32 price = p->opt[cur].price; - best = cur; - for (j = cur + 1; j <= last; j++) - { - UInt32 price2 = p->opt[j].price; - if (price >= price2) - { - price = price2; - best = j; - } - } - { - unsigned delta = best - cur; - if (delta != 0) - { - MOVE_POS(p, delta); - } - } - cur = best; - break; - } - - newLen = ReadMatchDistances(p, &numPairs); - - if (newLen >= p->numFastBytes) - { - p->numPairs = numPairs; - p->longestMatchLen = newLen; - break; - } - - curOpt = &p->opt[cur]; - - position++; - - // we need that check here, if skip_items in p->opt are possible - /* - if (curOpt->price >= kInfinityPrice) - continue; - */ - - prev = cur - curOpt->len; - - if (curOpt->len == 1) - { - state = (unsigned)p->opt[prev].state; - if (IsShortRep(curOpt)) - state = kShortRepNextStates[state]; - else - state = kLiteralNextStates[state]; - } - else - { - const COptimal *prevOpt; - UInt32 b0; - UInt32 dist = curOpt->dist; - - if (curOpt->extra) - { - prev -= (unsigned)curOpt->extra; - state = kState_RepAfterLit; - if (curOpt->extra == 1) - state = (dist < LZMA_NUM_REPS ? kState_RepAfterLit : kState_MatchAfterLit); - } - else - { - state = (unsigned)p->opt[prev].state; - if (dist < LZMA_NUM_REPS) - state = kRepNextStates[state]; - else - state = kMatchNextStates[state]; - } - - prevOpt = &p->opt[prev]; - b0 = prevOpt->reps[0]; - - if (dist < LZMA_NUM_REPS) - { - if (dist == 0) - { - reps[0] = b0; - reps[1] = prevOpt->reps[1]; - reps[2] = prevOpt->reps[2]; - reps[3] = prevOpt->reps[3]; - } - else - { - reps[1] = b0; - b0 = prevOpt->reps[1]; - if (dist == 1) - { - reps[0] = b0; - reps[2] = prevOpt->reps[2]; - reps[3] = prevOpt->reps[3]; - } - else - { - reps[2] = b0; - reps[0] = prevOpt->reps[dist]; - reps[3] = prevOpt->reps[dist ^ 1]; - } - } - } - else - { - reps[0] = (dist - LZMA_NUM_REPS + 1); - reps[1] = b0; - reps[2] = prevOpt->reps[1]; - reps[3] = prevOpt->reps[2]; - } - } - - curOpt->state = (CState)state; - curOpt->reps[0] = reps[0]; - curOpt->reps[1] = reps[1]; - curOpt->reps[2] = reps[2]; - curOpt->reps[3] = reps[3]; - - data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; - curByte = *data; - matchByte = *(data - reps[0]); - - posState = (position & p->pbMask); - - /* - The order of Price checks: - < LIT - <= SHORT_REP - < LIT : REP_0 - < REP [ : LIT : REP_0 ] - < MATCH [ : LIT : REP_0 ] - */ - - { - UInt32 curPrice = curOpt->price; - unsigned prob = p->isMatch[state][posState]; - matchPrice = curPrice + GET_PRICE_1(prob); - litPrice = curPrice + GET_PRICE_0(prob); - } - - nextOpt = &p->opt[(size_t)cur + 1]; - nextIsLit = False; - - // here we can allow skip_items in p->opt, if we don't check (nextOpt->price < kInfinityPrice) - // 18.new.06 - if ((nextOpt->price < kInfinityPrice - // && !IsLitState(state) - && matchByte == curByte) - || litPrice > nextOpt->price - ) - litPrice = 0; - else - { - const CLzmaProb *probs = LIT_PROBS(position, *(data - 1)); - litPrice += (!IsLitState(state) ? - LitEnc_Matched_GetPrice(probs, curByte, matchByte, p->ProbPrices) : - LitEnc_GetPrice(probs, curByte, p->ProbPrices)); - - if (litPrice < nextOpt->price) - { - nextOpt->price = litPrice; - nextOpt->len = 1; - MakeAs_Lit(nextOpt); - nextIsLit = True; - } - } - - repMatchPrice = matchPrice + GET_PRICE_1(p->isRep[state]); - - numAvailFull = p->numAvail; - { - unsigned temp = kNumOpts - 1 - cur; - if (numAvailFull > temp) - numAvailFull = (UInt32)temp; - } - - // 18.06 - // ---------- SHORT_REP ---------- - if (IsLitState(state)) // 18.new - if (matchByte == curByte) - if (repMatchPrice < nextOpt->price) // 18.new - // if (numAvailFull < 2 || data[1] != *(data - reps[0] + 1)) - if ( - // nextOpt->price >= kInfinityPrice || - nextOpt->len < 2 // we can check nextOpt->len, if skip items are not allowed in p->opt - || (nextOpt->dist != 0 - // && nextOpt->extra <= 1 // 17.old - ) - ) - { - UInt32 shortRepPrice = repMatchPrice + GetPrice_ShortRep(p, state, posState); - // if (shortRepPrice <= nextOpt->price) // 17.old - if (shortRepPrice < nextOpt->price) // 18.new - { - nextOpt->price = shortRepPrice; - nextOpt->len = 1; - MakeAs_ShortRep(nextOpt); - nextIsLit = False; - } - } - - if (numAvailFull < 2) - continue; - numAvail = (numAvailFull <= p->numFastBytes ? numAvailFull : p->numFastBytes); - - // numAvail <= p->numFastBytes - - // ---------- LIT : REP_0 ---------- - - if (!nextIsLit - && litPrice != 0 // 18.new - && matchByte != curByte - && numAvailFull > 2) - { - const Byte *data2 = data - reps[0]; - if (data[1] == data2[1] && data[2] == data2[2]) - { - unsigned len; - unsigned limit = p->numFastBytes + 1; - if (limit > numAvailFull) - limit = numAvailFull; - for (len = 3; len < limit && data[len] == data2[len]; len++) - {} - - { - unsigned state2 = kLiteralNextStates[state]; - unsigned posState2 = (position + 1) & p->pbMask; - UInt32 price = litPrice + GetPrice_Rep_0(p, state2, posState2); - { - unsigned offset = cur + len; - - if (last < offset) - last = offset; - - // do - { - UInt32 price2; - COptimal *opt; - len--; - // price2 = price + GetPrice_Len_Rep_0(p, len, state2, posState2); - price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState2, len); - - opt = &p->opt[offset]; - // offset--; - if (price2 < opt->price) - { - opt->price = price2; - opt->len = (UInt32)len; - opt->dist = 0; - opt->extra = 1; - } - } - // while (len >= 3); - } - } - } - } - - startLen = 2; /* speed optimization */ - - { - // ---------- REP ---------- - unsigned repIndex = 0; // 17.old - // unsigned repIndex = IsLitState(state) ? 0 : 1; // 18.notused - for (; repIndex < LZMA_NUM_REPS; repIndex++) - { - unsigned len; - UInt32 price; - const Byte *data2 = data - reps[repIndex]; - if (data[0] != data2[0] || data[1] != data2[1]) - continue; - - for (len = 2; len < numAvail && data[len] == data2[len]; len++) - {} - - // if (len < startLen) continue; // 18.new: speed optimization - - { - unsigned offset = cur + len; - if (last < offset) - last = offset; - } - { - unsigned len2 = len; - price = repMatchPrice + GetPrice_PureRep(p, repIndex, state, posState); - do - { - UInt32 price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState, len2); - COptimal *opt = &p->opt[cur + len2]; - if (price2 < opt->price) - { - opt->price = price2; - opt->len = (UInt32)len2; - opt->dist = (UInt32)repIndex; - opt->extra = 0; - } - } - while (--len2 >= 2); - } - - if (repIndex == 0) startLen = len + 1; // 17.old - // startLen = len + 1; // 18.new - - /* if (_maxMode) */ - { - // ---------- REP : LIT : REP_0 ---------- - // numFastBytes + 1 + numFastBytes - - unsigned len2 = len + 1; - unsigned limit = len2 + p->numFastBytes; - if (limit > numAvailFull) - limit = numAvailFull; - - len2 += 2; - if (len2 <= limit) - if (data[len2 - 2] == data2[len2 - 2]) - if (data[len2 - 1] == data2[len2 - 1]) - { - unsigned state2 = kRepNextStates[state]; - unsigned posState2 = (position + len) & p->pbMask; - price += GET_PRICE_LEN(&p->repLenEnc, posState, len) - + GET_PRICE_0(p->isMatch[state2][posState2]) - + LitEnc_Matched_GetPrice(LIT_PROBS(position + len, data[(size_t)len - 1]), - data[len], data2[len], p->ProbPrices); - - // state2 = kLiteralNextStates[state2]; - state2 = kState_LitAfterRep; - posState2 = (posState2 + 1) & p->pbMask; - - - price += GetPrice_Rep_0(p, state2, posState2); - - for (; len2 < limit && data[len2] == data2[len2]; len2++) - {} - - len2 -= len; - // if (len2 >= 3) - { - { - unsigned offset = cur + len + len2; - - if (last < offset) - last = offset; - // do - { - UInt32 price2; - COptimal *opt; - len2--; - // price2 = price + GetPrice_Len_Rep_0(p, len2, state2, posState2); - price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState2, len2); - - opt = &p->opt[offset]; - // offset--; - if (price2 < opt->price) - { - opt->price = price2; - opt->len = (UInt32)len2; - opt->extra = (CExtra)(len + 1); - opt->dist = (UInt32)repIndex; - } - } - // while (len2 >= 3); - } - } - } - } - } - } - - - // ---------- MATCH ---------- - /* for (unsigned len = 2; len <= newLen; len++) */ - if (newLen > numAvail) - { - newLen = numAvail; - for (numPairs = 0; newLen > MATCHES[numPairs]; numPairs += 2); - MATCHES[numPairs] = (UInt32)newLen; - numPairs += 2; - } - - // startLen = 2; /* speed optimization */ - - if (newLen >= startLen) - { - UInt32 normalMatchPrice = matchPrice + GET_PRICE_0(p->isRep[state]); - UInt32 dist; - unsigned offs, posSlot, len; - - { - unsigned offset = cur + newLen; - if (last < offset) - last = offset; - } - - offs = 0; - while (startLen > MATCHES[offs]) - offs += 2; - dist = MATCHES[(size_t)offs + 1]; - - // if (dist >= kNumFullDistances) - GetPosSlot2(dist, posSlot); - - for (len = /*2*/ startLen; ; len++) - { - UInt32 price = normalMatchPrice + GET_PRICE_LEN(&p->lenEnc, posState, len); - { - COptimal *opt; - unsigned lenNorm = len - 2; - lenNorm = GetLenToPosState2(lenNorm); - if (dist < kNumFullDistances) - price += p->distancesPrices[lenNorm][dist & (kNumFullDistances - 1)]; - else - price += p->posSlotPrices[lenNorm][posSlot] + p->alignPrices[dist & kAlignMask]; - - opt = &p->opt[cur + len]; - if (price < opt->price) - { - opt->price = price; - opt->len = (UInt32)len; - opt->dist = dist + LZMA_NUM_REPS; - opt->extra = 0; - } - } - - if (len == MATCHES[offs]) - { - // if (p->_maxMode) { - // MATCH : LIT : REP_0 - - const Byte *data2 = data - dist - 1; - unsigned len2 = len + 1; - unsigned limit = len2 + p->numFastBytes; - if (limit > numAvailFull) - limit = numAvailFull; - - len2 += 2; - if (len2 <= limit) - if (data[len2 - 2] == data2[len2 - 2]) - if (data[len2 - 1] == data2[len2 - 1]) - { - for (; len2 < limit && data[len2] == data2[len2]; len2++) - {} - - len2 -= len; - - // if (len2 >= 3) - { - unsigned state2 = kMatchNextStates[state]; - unsigned posState2 = (position + len) & p->pbMask; - unsigned offset; - price += GET_PRICE_0(p->isMatch[state2][posState2]); - price += LitEnc_Matched_GetPrice(LIT_PROBS(position + len, data[(size_t)len - 1]), - data[len], data2[len], p->ProbPrices); - - // state2 = kLiteralNextStates[state2]; - state2 = kState_LitAfterMatch; - - posState2 = (posState2 + 1) & p->pbMask; - price += GetPrice_Rep_0(p, state2, posState2); - - offset = cur + len + len2; - - if (last < offset) - last = offset; - // do - { - UInt32 price2; - COptimal *opt; - len2--; - // price2 = price + GetPrice_Len_Rep_0(p, len2, state2, posState2); - price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState2, len2); - opt = &p->opt[offset]; - // offset--; - if (price2 < opt->price) - { - opt->price = price2; - opt->len = (UInt32)len2; - opt->extra = (CExtra)(len + 1); - opt->dist = dist + LZMA_NUM_REPS; - } - } - // while (len2 >= 3); - } - - } - - offs += 2; - if (offs == numPairs) - break; - dist = MATCHES[(size_t)offs + 1]; - // if (dist >= kNumFullDistances) - GetPosSlot2(dist, posSlot); - } - } - } - } - - do - p->opt[last].price = kInfinityPrice; - while (--last); - - return Backward(p, cur); -} - - - -#define ChangePair(smallDist, bigDist) (((bigDist) >> 7) > (smallDist)) - - - -static unsigned GetOptimumFast(CLzmaEnc *p) -{ - UInt32 numAvail, mainDist; - unsigned mainLen, numPairs, repIndex, repLen, i; - const Byte *data; - - if (p->additionalOffset == 0) - mainLen = ReadMatchDistances(p, &numPairs); - else - { - mainLen = p->longestMatchLen; - numPairs = p->numPairs; - } - - numAvail = p->numAvail; - p->backRes = MARK_LIT; - if (numAvail < 2) - return 1; - // if (mainLen < 2 && p->state == 0) return 1; // 18.06.notused - if (numAvail > LZMA_MATCH_LEN_MAX) - numAvail = LZMA_MATCH_LEN_MAX; - data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; - repLen = repIndex = 0; - - for (i = 0; i < LZMA_NUM_REPS; i++) - { - unsigned len; - const Byte *data2 = data - p->reps[i]; - if (data[0] != data2[0] || data[1] != data2[1]) - continue; - for (len = 2; len < numAvail && data[len] == data2[len]; len++) - {} - if (len >= p->numFastBytes) - { - p->backRes = (UInt32)i; - MOVE_POS(p, len - 1) - return len; - } - if (len > repLen) - { - repIndex = i; - repLen = len; - } - } - - if (mainLen >= p->numFastBytes) - { - p->backRes = p->matches[(size_t)numPairs - 1] + LZMA_NUM_REPS; - MOVE_POS(p, mainLen - 1) - return mainLen; - } - - mainDist = 0; /* for GCC */ - - if (mainLen >= 2) - { - mainDist = p->matches[(size_t)numPairs - 1]; - while (numPairs > 2) - { - UInt32 dist2; - if (mainLen != p->matches[(size_t)numPairs - 4] + 1) - break; - dist2 = p->matches[(size_t)numPairs - 3]; - if (!ChangePair(dist2, mainDist)) - break; - numPairs -= 2; - mainLen--; - mainDist = dist2; - } - if (mainLen == 2 && mainDist >= 0x80) - mainLen = 1; - } - - if (repLen >= 2) - if ( repLen + 1 >= mainLen - || (repLen + 2 >= mainLen && mainDist >= (1 << 9)) - || (repLen + 3 >= mainLen && mainDist >= (1 << 15))) - { - p->backRes = (UInt32)repIndex; - MOVE_POS(p, repLen - 1) - return repLen; - } - - if (mainLen < 2 || numAvail <= 2) - return 1; - - { - unsigned len1 = ReadMatchDistances(p, &p->numPairs); - p->longestMatchLen = len1; - - if (len1 >= 2) - { - UInt32 newDist = p->matches[(size_t)p->numPairs - 1]; - if ( (len1 >= mainLen && newDist < mainDist) - || (len1 == mainLen + 1 && !ChangePair(mainDist, newDist)) - || (len1 > mainLen + 1) - || (len1 + 1 >= mainLen && mainLen >= 3 && ChangePair(newDist, mainDist))) - return 1; - } - } - - data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; - - for (i = 0; i < LZMA_NUM_REPS; i++) - { - unsigned len, limit; - const Byte *data2 = data - p->reps[i]; - if (data[0] != data2[0] || data[1] != data2[1]) - continue; - limit = mainLen - 1; - for (len = 2;; len++) - { - if (len >= limit) - return 1; - if (data[len] != data2[len]) - break; - } - } - - p->backRes = mainDist + LZMA_NUM_REPS; - if (mainLen != 2) - { - MOVE_POS(p, mainLen - 2) - } - return mainLen; -} - - - - -static void WriteEndMarker(CLzmaEnc *p, unsigned posState) -{ - UInt32 range; - range = p->rc.range; - { - UInt32 ttt, newBound; - CLzmaProb *prob = &p->isMatch[p->state][posState]; - RC_BIT_PRE(&p->rc, prob) - RC_BIT_1(&p->rc, prob) - prob = &p->isRep[p->state]; - RC_BIT_PRE(&p->rc, prob) - RC_BIT_0(&p->rc, prob) - } - p->state = kMatchNextStates[p->state]; - - p->rc.range = range; - LenEnc_Encode(&p->lenProbs, &p->rc, 0, posState); - range = p->rc.range; - - { - // RcTree_Encode_PosSlot(&p->rc, p->posSlotEncoder[0], (1 << kNumPosSlotBits) - 1); - CLzmaProb *probs = p->posSlotEncoder[0]; - unsigned m = 1; - do - { - UInt32 ttt, newBound; - RC_BIT_PRE(p, probs + m) - RC_BIT_1(&p->rc, probs + m); - m = (m << 1) + 1; - } - while (m < (1 << kNumPosSlotBits)); - } - { - // RangeEnc_EncodeDirectBits(&p->rc, ((UInt32)1 << (30 - kNumAlignBits)) - 1, 30 - kNumAlignBits); UInt32 range = p->range; - unsigned numBits = 30 - kNumAlignBits; - do - { - range >>= 1; - p->rc.low += range; - RC_NORM(&p->rc) - } - while (--numBits); - } - - { - // RcTree_ReverseEncode(&p->rc, p->posAlignEncoder, kNumAlignBits, kAlignMask); - CLzmaProb *probs = p->posAlignEncoder; - unsigned m = 1; - do - { - UInt32 ttt, newBound; - RC_BIT_PRE(p, probs + m) - RC_BIT_1(&p->rc, probs + m); - m = (m << 1) + 1; - } - while (m < kAlignTableSize); - } - p->rc.range = range; -} - - -static SRes CheckErrors(CLzmaEnc *p) -{ - if (p->result != SZ_OK) - return p->result; - if (p->rc.res != SZ_OK) - p->result = SZ_ERROR_WRITE; - - #ifndef _7ZIP_ST - if ( - // p->mf_Failure || - (p->mtMode && - ( // p->matchFinderMt.failure_LZ_LZ || - p->matchFinderMt.failure_LZ_BT)) - ) - { - p->result = MY_HRES_ERROR__INTERNAL_ERROR; - // printf("\nCheckErrors p->matchFinderMt.failureLZ\n"); - } - #endif - - if (MFB.result != SZ_OK) - p->result = SZ_ERROR_READ; - - if (p->result != SZ_OK) - p->finished = True; - return p->result; -} - - -MY_NO_INLINE static SRes Flush(CLzmaEnc *p, UInt32 nowPos) -{ - /* ReleaseMFStream(); */ - p->finished = True; - if (p->writeEndMark) - WriteEndMarker(p, nowPos & p->pbMask); - RangeEnc_FlushData(&p->rc); - RangeEnc_FlushStream(&p->rc); - return CheckErrors(p); -} - - -MY_NO_INLINE static void FillAlignPrices(CLzmaEnc *p) -{ - unsigned i; - const CProbPrice *ProbPrices = p->ProbPrices; - const CLzmaProb *probs = p->posAlignEncoder; - // p->alignPriceCount = 0; - for (i = 0; i < kAlignTableSize / 2; i++) - { - UInt32 price = 0; - unsigned sym = i; - unsigned m = 1; - unsigned bit; - UInt32 prob; - bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[m], bit); m = (m << 1) + bit; - bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[m], bit); m = (m << 1) + bit; - bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[m], bit); m = (m << 1) + bit; - prob = probs[m]; - p->alignPrices[i ] = price + GET_PRICEa_0(prob); - p->alignPrices[i + 8] = price + GET_PRICEa_1(prob); - // p->alignPrices[i] = RcTree_ReverseGetPrice(p->posAlignEncoder, kNumAlignBits, i, p->ProbPrices); - } -} - - -MY_NO_INLINE static void FillDistancesPrices(CLzmaEnc *p) -{ - // int y; for (y = 0; y < 100; y++) { - - UInt32 tempPrices[kNumFullDistances]; - unsigned i, lps; - - const CProbPrice *ProbPrices = p->ProbPrices; - p->matchPriceCount = 0; - - for (i = kStartPosModelIndex / 2; i < kNumFullDistances / 2; i++) - { - unsigned posSlot = GetPosSlot1(i); - unsigned footerBits = (posSlot >> 1) - 1; - unsigned base = ((2 | (posSlot & 1)) << footerBits); - const CLzmaProb *probs = p->posEncoders + (size_t)base * 2; - // tempPrices[i] = RcTree_ReverseGetPrice(p->posEncoders + base, footerBits, i - base, p->ProbPrices); - UInt32 price = 0; - unsigned m = 1; - unsigned sym = i; - unsigned offset = (unsigned)1 << footerBits; - base += i; - - if (footerBits) - do - { - unsigned bit = sym & 1; - sym >>= 1; - price += GET_PRICEa(probs[m], bit); - m = (m << 1) + bit; - } - while (--footerBits); - - { - unsigned prob = probs[m]; - tempPrices[base ] = price + GET_PRICEa_0(prob); - tempPrices[base + offset] = price + GET_PRICEa_1(prob); - } - } - - for (lps = 0; lps < kNumLenToPosStates; lps++) - { - unsigned slot; - unsigned distTableSize2 = (p->distTableSize + 1) >> 1; - UInt32 *posSlotPrices = p->posSlotPrices[lps]; - const CLzmaProb *probs = p->posSlotEncoder[lps]; - - for (slot = 0; slot < distTableSize2; slot++) - { - // posSlotPrices[slot] = RcTree_GetPrice(encoder, kNumPosSlotBits, slot, p->ProbPrices); - UInt32 price; - unsigned bit; - unsigned sym = slot + (1 << (kNumPosSlotBits - 1)); - unsigned prob; - bit = sym & 1; sym >>= 1; price = GET_PRICEa(probs[sym], bit); - bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); - bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); - bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); - bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); - prob = probs[(size_t)slot + (1 << (kNumPosSlotBits - 1))]; - posSlotPrices[(size_t)slot * 2 ] = price + GET_PRICEa_0(prob); - posSlotPrices[(size_t)slot * 2 + 1] = price + GET_PRICEa_1(prob); - } - - { - UInt32 delta = ((UInt32)((kEndPosModelIndex / 2 - 1) - kNumAlignBits) << kNumBitPriceShiftBits); - for (slot = kEndPosModelIndex / 2; slot < distTableSize2; slot++) - { - posSlotPrices[(size_t)slot * 2 ] += delta; - posSlotPrices[(size_t)slot * 2 + 1] += delta; - delta += ((UInt32)1 << kNumBitPriceShiftBits); - } - } - - { - UInt32 *dp = p->distancesPrices[lps]; - - dp[0] = posSlotPrices[0]; - dp[1] = posSlotPrices[1]; - dp[2] = posSlotPrices[2]; - dp[3] = posSlotPrices[3]; - - for (i = 4; i < kNumFullDistances; i += 2) - { - UInt32 slotPrice = posSlotPrices[GetPosSlot1(i)]; - dp[i ] = slotPrice + tempPrices[i]; - dp[i + 1] = slotPrice + tempPrices[i + 1]; - } - } - } - // } -} - - - -static void LzmaEnc_Construct(CLzmaEnc *p) -{ - RangeEnc_Construct(&p->rc); - MatchFinder_Construct(&MFB); - - #ifndef _7ZIP_ST - p->matchFinderMt.MatchFinder = &MFB; - MatchFinderMt_Construct(&p->matchFinderMt); - #endif - - { - CLzmaEncProps props; - LzmaEncProps_Init(&props); - LzmaEnc_SetProps(p, &props); - } - - #ifndef LZMA_LOG_BSR - LzmaEnc_FastPosInit(p->g_FastPos); - #endif - - LzmaEnc_InitPriceTables(p->ProbPrices); - p->litProbs = NULL; - p->saveState.litProbs = NULL; -} - -CLzmaEncHandle LzmaEnc_Create(ISzAllocPtr alloc) -{ - void *p; - p = ISzAlloc_Alloc(alloc, sizeof(CLzmaEnc)); - if (p) - LzmaEnc_Construct((CLzmaEnc *)p); - return p; -} - -static void LzmaEnc_FreeLits(CLzmaEnc *p, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, p->litProbs); - ISzAlloc_Free(alloc, p->saveState.litProbs); - p->litProbs = NULL; - p->saveState.litProbs = NULL; -} - -static void LzmaEnc_Destruct(CLzmaEnc *p, ISzAllocPtr alloc, ISzAllocPtr allocBig) -{ - #ifndef _7ZIP_ST - MatchFinderMt_Destruct(&p->matchFinderMt, allocBig); - #endif - - MatchFinder_Free(&MFB, allocBig); - LzmaEnc_FreeLits(p, alloc); - RangeEnc_Free(&p->rc, alloc); -} - -void LzmaEnc_Destroy(CLzmaEncHandle p, ISzAllocPtr alloc, ISzAllocPtr allocBig) -{ - LzmaEnc_Destruct((CLzmaEnc *)p, alloc, allocBig); - ISzAlloc_Free(alloc, p); -} - - -MY_NO_INLINE -static SRes LzmaEnc_CodeOneBlock(CLzmaEnc *p, UInt32 maxPackSize, UInt32 maxUnpackSize) -{ - UInt32 nowPos32, startPos32; - if (p->needInit) - { - #ifndef _7ZIP_ST - if (p->mtMode) - { - RINOK(MatchFinderMt_InitMt(&p->matchFinderMt)); - } - #endif - p->matchFinder.Init(p->matchFinderObj); - p->needInit = 0; - } - - if (p->finished) - return p->result; - RINOK(CheckErrors(p)); - - nowPos32 = (UInt32)p->nowPos64; - startPos32 = nowPos32; - - if (p->nowPos64 == 0) - { - unsigned numPairs; - Byte curByte; - if (p->matchFinder.GetNumAvailableBytes(p->matchFinderObj) == 0) - return Flush(p, nowPos32); - ReadMatchDistances(p, &numPairs); - RangeEnc_EncodeBit_0(&p->rc, &p->isMatch[kState_Start][0]); - // p->state = kLiteralNextStates[p->state]; - curByte = *(p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - p->additionalOffset); - LitEnc_Encode(&p->rc, p->litProbs, curByte); - p->additionalOffset--; - nowPos32++; - } - - if (p->matchFinder.GetNumAvailableBytes(p->matchFinderObj) != 0) - - for (;;) - { - UInt32 dist; - unsigned len, posState; - UInt32 range, ttt, newBound; - CLzmaProb *probs; - - if (p->fastMode) - len = GetOptimumFast(p); - else - { - unsigned oci = p->optCur; - if (p->optEnd == oci) - len = GetOptimum(p, nowPos32); - else - { - const COptimal *opt = &p->opt[oci]; - len = opt->len; - p->backRes = opt->dist; - p->optCur = oci + 1; - } - } - - posState = (unsigned)nowPos32 & p->pbMask; - range = p->rc.range; - probs = &p->isMatch[p->state][posState]; - - RC_BIT_PRE(&p->rc, probs) - - dist = p->backRes; - - #ifdef SHOW_STAT2 - printf("\n pos = %6X, len = %3u pos = %6u", nowPos32, len, dist); - #endif - - if (dist == MARK_LIT) - { - Byte curByte; - const Byte *data; - unsigned state; - - RC_BIT_0(&p->rc, probs); - p->rc.range = range; - data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - p->additionalOffset; - probs = LIT_PROBS(nowPos32, *(data - 1)); - curByte = *data; - state = p->state; - p->state = kLiteralNextStates[state]; - if (IsLitState(state)) - LitEnc_Encode(&p->rc, probs, curByte); - else - LitEnc_EncodeMatched(&p->rc, probs, curByte, *(data - p->reps[0])); - } - else - { - RC_BIT_1(&p->rc, probs); - probs = &p->isRep[p->state]; - RC_BIT_PRE(&p->rc, probs) - - if (dist < LZMA_NUM_REPS) - { - RC_BIT_1(&p->rc, probs); - probs = &p->isRepG0[p->state]; - RC_BIT_PRE(&p->rc, probs) - if (dist == 0) - { - RC_BIT_0(&p->rc, probs); - probs = &p->isRep0Long[p->state][posState]; - RC_BIT_PRE(&p->rc, probs) - if (len != 1) - { - RC_BIT_1_BASE(&p->rc, probs); - } - else - { - RC_BIT_0_BASE(&p->rc, probs); - p->state = kShortRepNextStates[p->state]; - } - } - else - { - RC_BIT_1(&p->rc, probs); - probs = &p->isRepG1[p->state]; - RC_BIT_PRE(&p->rc, probs) - if (dist == 1) - { - RC_BIT_0_BASE(&p->rc, probs); - dist = p->reps[1]; - } - else - { - RC_BIT_1(&p->rc, probs); - probs = &p->isRepG2[p->state]; - RC_BIT_PRE(&p->rc, probs) - if (dist == 2) - { - RC_BIT_0_BASE(&p->rc, probs); - dist = p->reps[2]; - } - else - { - RC_BIT_1_BASE(&p->rc, probs); - dist = p->reps[3]; - p->reps[3] = p->reps[2]; - } - p->reps[2] = p->reps[1]; - } - p->reps[1] = p->reps[0]; - p->reps[0] = dist; - } - - RC_NORM(&p->rc) - - p->rc.range = range; - - if (len != 1) - { - LenEnc_Encode(&p->repLenProbs, &p->rc, len - LZMA_MATCH_LEN_MIN, posState); - --p->repLenEncCounter; - p->state = kRepNextStates[p->state]; - } - } - else - { - unsigned posSlot; - RC_BIT_0(&p->rc, probs); - p->rc.range = range; - p->state = kMatchNextStates[p->state]; - - LenEnc_Encode(&p->lenProbs, &p->rc, len - LZMA_MATCH_LEN_MIN, posState); - // --p->lenEnc.counter; - - dist -= LZMA_NUM_REPS; - p->reps[3] = p->reps[2]; - p->reps[2] = p->reps[1]; - p->reps[1] = p->reps[0]; - p->reps[0] = dist + 1; - - p->matchPriceCount++; - GetPosSlot(dist, posSlot); - // RcTree_Encode_PosSlot(&p->rc, p->posSlotEncoder[GetLenToPosState(len)], posSlot); - { - UInt32 sym = (UInt32)posSlot + (1 << kNumPosSlotBits); - range = p->rc.range; - probs = p->posSlotEncoder[GetLenToPosState(len)]; - do - { - CLzmaProb *prob = probs + (sym >> kNumPosSlotBits); - UInt32 bit = (sym >> (kNumPosSlotBits - 1)) & 1; - sym <<= 1; - RC_BIT(&p->rc, prob, bit); - } - while (sym < (1 << kNumPosSlotBits * 2)); - p->rc.range = range; - } - - if (dist >= kStartPosModelIndex) - { - unsigned footerBits = ((posSlot >> 1) - 1); - - if (dist < kNumFullDistances) - { - unsigned base = ((2 | (posSlot & 1)) << footerBits); - RcTree_ReverseEncode(&p->rc, p->posEncoders + base, footerBits, (unsigned)(dist /* - base */)); - } - else - { - UInt32 pos2 = (dist | 0xF) << (32 - footerBits); - range = p->rc.range; - // RangeEnc_EncodeDirectBits(&p->rc, posReduced >> kNumAlignBits, footerBits - kNumAlignBits); - /* - do - { - range >>= 1; - p->rc.low += range & (0 - ((dist >> --footerBits) & 1)); - RC_NORM(&p->rc) - } - while (footerBits > kNumAlignBits); - */ - do - { - range >>= 1; - p->rc.low += range & (0 - (pos2 >> 31)); - pos2 += pos2; - RC_NORM(&p->rc) - } - while (pos2 != 0xF0000000); - - - // RcTree_ReverseEncode(&p->rc, p->posAlignEncoder, kNumAlignBits, posReduced & kAlignMask); - - { - unsigned m = 1; - unsigned bit; - bit = dist & 1; dist >>= 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); m = (m << 1) + bit; - bit = dist & 1; dist >>= 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); m = (m << 1) + bit; - bit = dist & 1; dist >>= 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); m = (m << 1) + bit; - bit = dist & 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); - p->rc.range = range; - // p->alignPriceCount++; - } - } - } - } - } - - nowPos32 += (UInt32)len; - p->additionalOffset -= len; - - if (p->additionalOffset == 0) - { - UInt32 processed; - - if (!p->fastMode) - { - /* - if (p->alignPriceCount >= 16) // kAlignTableSize - FillAlignPrices(p); - if (p->matchPriceCount >= 128) - FillDistancesPrices(p); - if (p->lenEnc.counter <= 0) - LenPriceEnc_UpdateTables(&p->lenEnc, 1 << p->pb, &p->lenProbs, p->ProbPrices); - */ - if (p->matchPriceCount >= 64) - { - FillAlignPrices(p); - // { int y; for (y = 0; y < 100; y++) { - FillDistancesPrices(p); - // }} - LenPriceEnc_UpdateTables(&p->lenEnc, (unsigned)1 << p->pb, &p->lenProbs, p->ProbPrices); - } - if (p->repLenEncCounter <= 0) - { - p->repLenEncCounter = REP_LEN_COUNT; - LenPriceEnc_UpdateTables(&p->repLenEnc, (unsigned)1 << p->pb, &p->repLenProbs, p->ProbPrices); - } - } - - if (p->matchFinder.GetNumAvailableBytes(p->matchFinderObj) == 0) - break; - processed = nowPos32 - startPos32; - - if (maxPackSize) - { - if (processed + kNumOpts + 300 >= maxUnpackSize - || RangeEnc_GetProcessed_sizet(&p->rc) + kPackReserve >= maxPackSize) - break; - } - else if (processed >= (1 << 17)) - { - p->nowPos64 += nowPos32 - startPos32; - return CheckErrors(p); - } - } - } - - p->nowPos64 += nowPos32 - startPos32; - return Flush(p, nowPos32); -} - - - -#define kBigHashDicLimit ((UInt32)1 << 24) - -static SRes LzmaEnc_Alloc(CLzmaEnc *p, UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig) -{ - UInt32 beforeSize = kNumOpts; - UInt32 dictSize; - - if (!RangeEnc_Alloc(&p->rc, alloc)) - return SZ_ERROR_MEM; - - #ifndef _7ZIP_ST - p->mtMode = (p->multiThread && !p->fastMode && (MFB.btMode != 0)); - #endif - - { - unsigned lclp = p->lc + p->lp; - if (!p->litProbs || !p->saveState.litProbs || p->lclp != lclp) - { - LzmaEnc_FreeLits(p, alloc); - p->litProbs = (CLzmaProb *)ISzAlloc_Alloc(alloc, ((UInt32)0x300 << lclp) * sizeof(CLzmaProb)); - p->saveState.litProbs = (CLzmaProb *)ISzAlloc_Alloc(alloc, ((UInt32)0x300 << lclp) * sizeof(CLzmaProb)); - if (!p->litProbs || !p->saveState.litProbs) - { - LzmaEnc_FreeLits(p, alloc); - return SZ_ERROR_MEM; - } - p->lclp = lclp; - } - } - - MFB.bigHash = (Byte)(p->dictSize > kBigHashDicLimit ? 1 : 0); - - - dictSize = p->dictSize; - if (dictSize == ((UInt32)2 << 30) || - dictSize == ((UInt32)3 << 30)) - { - /* 21.03 : here we reduce the dictionary for 2 reasons: - 1) we don't want 32-bit back_distance matches in decoder for 2 GB dictionary. - 2) we want to elimate useless last MatchFinder_Normalize3() for corner cases, - where data size is aligned for 1 GB: 5/6/8 GB. - That reducing must be >= 1 for such corner cases. */ - dictSize -= 1; - } - - if (beforeSize + dictSize < keepWindowSize) - beforeSize = keepWindowSize - dictSize; - - /* in worst case we can look ahead for - max(LZMA_MATCH_LEN_MAX, numFastBytes + 1 + numFastBytes) bytes. - we send larger value for (keepAfter) to MantchFinder_Create(): - (numFastBytes + LZMA_MATCH_LEN_MAX + 1) - */ - - #ifndef _7ZIP_ST - if (p->mtMode) - { - RINOK(MatchFinderMt_Create(&p->matchFinderMt, dictSize, beforeSize, - p->numFastBytes, LZMA_MATCH_LEN_MAX + 1 /* 18.04 */ - , allocBig)); - p->matchFinderObj = &p->matchFinderMt; - MFB.bigHash = (Byte)( - (p->dictSize > kBigHashDicLimit && MFB.hashMask >= 0xFFFFFF) ? 1 : 0); - MatchFinderMt_CreateVTable(&p->matchFinderMt, &p->matchFinder); - } - else - #endif - { - if (!MatchFinder_Create(&MFB, dictSize, beforeSize, - p->numFastBytes, LZMA_MATCH_LEN_MAX + 1 /* 21.03 */ - , allocBig)) - return SZ_ERROR_MEM; - p->matchFinderObj = &MFB; - MatchFinder_CreateVTable(&MFB, &p->matchFinder); - } - - return SZ_OK; -} - -static void LzmaEnc_Init(CLzmaEnc *p) -{ - unsigned i; - p->state = 0; - p->reps[0] = - p->reps[1] = - p->reps[2] = - p->reps[3] = 1; - - RangeEnc_Init(&p->rc); - - for (i = 0; i < (1 << kNumAlignBits); i++) - p->posAlignEncoder[i] = kProbInitValue; - - for (i = 0; i < kNumStates; i++) - { - unsigned j; - for (j = 0; j < LZMA_NUM_PB_STATES_MAX; j++) - { - p->isMatch[i][j] = kProbInitValue; - p->isRep0Long[i][j] = kProbInitValue; - } - p->isRep[i] = kProbInitValue; - p->isRepG0[i] = kProbInitValue; - p->isRepG1[i] = kProbInitValue; - p->isRepG2[i] = kProbInitValue; - } - - { - for (i = 0; i < kNumLenToPosStates; i++) - { - CLzmaProb *probs = p->posSlotEncoder[i]; - unsigned j; - for (j = 0; j < (1 << kNumPosSlotBits); j++) - probs[j] = kProbInitValue; - } - } - { - for (i = 0; i < kNumFullDistances; i++) - p->posEncoders[i] = kProbInitValue; - } - - { - UInt32 num = (UInt32)0x300 << (p->lp + p->lc); - UInt32 k; - CLzmaProb *probs = p->litProbs; - for (k = 0; k < num; k++) - probs[k] = kProbInitValue; - } - - - LenEnc_Init(&p->lenProbs); - LenEnc_Init(&p->repLenProbs); - - p->optEnd = 0; - p->optCur = 0; - - { - for (i = 0; i < kNumOpts; i++) - p->opt[i].price = kInfinityPrice; - } - - p->additionalOffset = 0; - - p->pbMask = ((unsigned)1 << p->pb) - 1; - p->lpMask = ((UInt32)0x100 << p->lp) - ((unsigned)0x100 >> p->lc); - - // p->mf_Failure = False; -} - - -static void LzmaEnc_InitPrices(CLzmaEnc *p) -{ - if (!p->fastMode) - { - FillDistancesPrices(p); - FillAlignPrices(p); - } - - p->lenEnc.tableSize = - p->repLenEnc.tableSize = - p->numFastBytes + 1 - LZMA_MATCH_LEN_MIN; - - p->repLenEncCounter = REP_LEN_COUNT; - - LenPriceEnc_UpdateTables(&p->lenEnc, (unsigned)1 << p->pb, &p->lenProbs, p->ProbPrices); - LenPriceEnc_UpdateTables(&p->repLenEnc, (unsigned)1 << p->pb, &p->repLenProbs, p->ProbPrices); -} - -static SRes LzmaEnc_AllocAndInit(CLzmaEnc *p, UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig) -{ - unsigned i; - for (i = kEndPosModelIndex / 2; i < kDicLogSizeMax; i++) - if (p->dictSize <= ((UInt32)1 << i)) - break; - p->distTableSize = i * 2; - - p->finished = False; - p->result = SZ_OK; - RINOK(LzmaEnc_Alloc(p, keepWindowSize, alloc, allocBig)); - LzmaEnc_Init(p); - LzmaEnc_InitPrices(p); - p->nowPos64 = 0; - return SZ_OK; -} - -static SRes LzmaEnc_Prepare(CLzmaEncHandle pp, ISeqOutStream *outStream, ISeqInStream *inStream, - ISzAllocPtr alloc, ISzAllocPtr allocBig) -{ - CLzmaEnc *p = (CLzmaEnc *)pp; - MFB.stream = inStream; - p->needInit = 1; - p->rc.outStream = outStream; - return LzmaEnc_AllocAndInit(p, 0, alloc, allocBig); -} - -SRes LzmaEnc_PrepareForLzma2(CLzmaEncHandle pp, - ISeqInStream *inStream, UInt32 keepWindowSize, - ISzAllocPtr alloc, ISzAllocPtr allocBig) -{ - CLzmaEnc *p = (CLzmaEnc *)pp; - MFB.stream = inStream; - p->needInit = 1; - return LzmaEnc_AllocAndInit(p, keepWindowSize, alloc, allocBig); -} - -static void LzmaEnc_SetInputBuf(CLzmaEnc *p, const Byte *src, SizeT srcLen) -{ - MFB.directInput = 1; - MFB.bufferBase = (Byte *)src; - MFB.directInputRem = srcLen; -} - -SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen, - UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig) -{ - CLzmaEnc *p = (CLzmaEnc *)pp; - LzmaEnc_SetInputBuf(p, src, srcLen); - p->needInit = 1; - - LzmaEnc_SetDataSize(pp, srcLen); - return LzmaEnc_AllocAndInit(p, keepWindowSize, alloc, allocBig); -} - -void LzmaEnc_Finish(CLzmaEncHandle pp) -{ - #ifndef _7ZIP_ST - CLzmaEnc *p = (CLzmaEnc *)pp; - if (p->mtMode) - MatchFinderMt_ReleaseStream(&p->matchFinderMt); - #else - UNUSED_VAR(pp); - #endif -} - - -typedef struct -{ - ISeqOutStream vt; - Byte *data; - SizeT rem; - BoolInt overflow; -} CLzmaEnc_SeqOutStreamBuf; - -static size_t SeqOutStreamBuf_Write(const ISeqOutStream *pp, const void *data, size_t size) -{ - CLzmaEnc_SeqOutStreamBuf *p = CONTAINER_FROM_VTBL(pp, CLzmaEnc_SeqOutStreamBuf, vt); - if (p->rem < size) - { - size = p->rem; - p->overflow = True; - } - if (size != 0) - { - memcpy(p->data, data, size); - p->rem -= size; - p->data += size; - } - return size; -} - - -/* -UInt32 LzmaEnc_GetNumAvailableBytes(CLzmaEncHandle pp) -{ - const CLzmaEnc *p = (CLzmaEnc *)pp; - return p->matchFinder.GetNumAvailableBytes(p->matchFinderObj); -} -*/ - -const Byte *LzmaEnc_GetCurBuf(CLzmaEncHandle pp) -{ - const CLzmaEnc *p = (CLzmaEnc *)pp; - return p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - p->additionalOffset; -} - - -// (desiredPackSize == 0) is not allowed -SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit, - Byte *dest, size_t *destLen, UInt32 desiredPackSize, UInt32 *unpackSize) -{ - CLzmaEnc *p = (CLzmaEnc *)pp; - UInt64 nowPos64; - SRes res; - CLzmaEnc_SeqOutStreamBuf outStream; - - outStream.vt.Write = SeqOutStreamBuf_Write; - outStream.data = dest; - outStream.rem = *destLen; - outStream.overflow = False; - - p->writeEndMark = False; - p->finished = False; - p->result = SZ_OK; - - if (reInit) - LzmaEnc_Init(p); - LzmaEnc_InitPrices(p); - RangeEnc_Init(&p->rc); - p->rc.outStream = &outStream.vt; - nowPos64 = p->nowPos64; - - res = LzmaEnc_CodeOneBlock(p, desiredPackSize, *unpackSize); - - *unpackSize = (UInt32)(p->nowPos64 - nowPos64); - *destLen -= outStream.rem; - if (outStream.overflow) - return SZ_ERROR_OUTPUT_EOF; - - return res; -} - - -MY_NO_INLINE -static SRes LzmaEnc_Encode2(CLzmaEnc *p, ICompressProgress *progress) -{ - SRes res = SZ_OK; - - #ifndef _7ZIP_ST - Byte allocaDummy[0x300]; - allocaDummy[0] = 0; - allocaDummy[1] = allocaDummy[0]; - #endif - - for (;;) - { - res = LzmaEnc_CodeOneBlock(p, 0, 0); - if (res != SZ_OK || p->finished) - break; - if (progress) - { - res = ICompressProgress_Progress(progress, p->nowPos64, RangeEnc_GetProcessed(&p->rc)); - if (res != SZ_OK) - { - res = SZ_ERROR_PROGRESS; - break; - } - } - } - - LzmaEnc_Finish(p); - - /* - if (res == SZ_OK && !Inline_MatchFinder_IsFinishedOK(&MFB)) - res = SZ_ERROR_FAIL; - } - */ - - return res; -} - - -SRes LzmaEnc_Encode(CLzmaEncHandle pp, ISeqOutStream *outStream, ISeqInStream *inStream, ICompressProgress *progress, - ISzAllocPtr alloc, ISzAllocPtr allocBig) -{ - RINOK(LzmaEnc_Prepare(pp, outStream, inStream, alloc, allocBig)); - return LzmaEnc_Encode2((CLzmaEnc *)pp, progress); -} - - -SRes LzmaEnc_WriteProperties(CLzmaEncHandle pp, Byte *props, SizeT *size) -{ - if (*size < LZMA_PROPS_SIZE) - return SZ_ERROR_PARAM; - *size = LZMA_PROPS_SIZE; - { - const CLzmaEnc *p = (const CLzmaEnc *)pp; - const UInt32 dictSize = p->dictSize; - UInt32 v; - props[0] = (Byte)((p->pb * 5 + p->lp) * 9 + p->lc); - - // we write aligned dictionary value to properties for lzma decoder - if (dictSize >= ((UInt32)1 << 21)) - { - const UInt32 kDictMask = ((UInt32)1 << 20) - 1; - v = (dictSize + kDictMask) & ~kDictMask; - if (v < dictSize) - v = dictSize; - } - else - { - unsigned i = 11 * 2; - do - { - v = (UInt32)(2 + (i & 1)) << (i >> 1); - i++; - } - while (v < dictSize); - } - - SetUi32(props + 1, v); - return SZ_OK; - } -} - - -unsigned LzmaEnc_IsWriteEndMark(CLzmaEncHandle pp) -{ - return (unsigned)((CLzmaEnc *)pp)->writeEndMark; -} - - -SRes LzmaEnc_MemEncode(CLzmaEncHandle pp, Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, - int writeEndMark, ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig) -{ - SRes res; - CLzmaEnc *p = (CLzmaEnc *)pp; - - CLzmaEnc_SeqOutStreamBuf outStream; - - outStream.vt.Write = SeqOutStreamBuf_Write; - outStream.data = dest; - outStream.rem = *destLen; - outStream.overflow = False; - - p->writeEndMark = writeEndMark; - p->rc.outStream = &outStream.vt; - - res = LzmaEnc_MemPrepare(pp, src, srcLen, 0, alloc, allocBig); - - if (res == SZ_OK) - { - res = LzmaEnc_Encode2(p, progress); - if (res == SZ_OK && p->nowPos64 != srcLen) - res = SZ_ERROR_FAIL; - } - - *destLen -= outStream.rem; - if (outStream.overflow) - return SZ_ERROR_OUTPUT_EOF; - return res; -} - - -SRes LzmaEncode(Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, - const CLzmaEncProps *props, Byte *propsEncoded, SizeT *propsSize, int writeEndMark, - ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig) -{ - CLzmaEnc *p = (CLzmaEnc *)LzmaEnc_Create(alloc); - SRes res; - if (!p) - return SZ_ERROR_MEM; - - res = LzmaEnc_SetProps(p, props); - if (res == SZ_OK) - { - res = LzmaEnc_WriteProperties(p, propsEncoded, propsSize); - if (res == SZ_OK) - res = LzmaEnc_MemEncode(p, dest, destLen, src, srcLen, - writeEndMark, progress, alloc, allocBig); - } - - LzmaEnc_Destroy(p, alloc, allocBig); - return res; -} - - -/* -#ifndef _7ZIP_ST -void LzmaEnc_GetLzThreads(CLzmaEncHandle pp, HANDLE lz_threads[2]) -{ - const CLzmaEnc *p = (CLzmaEnc *)pp; - lz_threads[0] = p->matchFinderMt.hashSync.thread; - lz_threads[1] = p->matchFinderMt.btSync.thread; -} -#endif -*/ +/* LzmaEnc.c -- LZMA Encoder +2022-07-15: Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +/* #define SHOW_STAT */ +/* #define SHOW_STAT2 */ + +#if defined(SHOW_STAT) || defined(SHOW_STAT2) +#include +#endif + +#include "CpuArch.h" +#include "LzmaEnc.h" + +#include "LzFind.h" +#ifndef _7ZIP_ST +#include "LzFindMt.h" +#endif + +/* the following LzmaEnc_* declarations is internal LZMA interface for LZMA2 encoder */ + +SRes LzmaEnc_PrepareForLzma2(CLzmaEncHandle pp, ISeqInStream *inStream, UInt32 keepWindowSize, + ISzAllocPtr alloc, ISzAllocPtr allocBig); +SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen, + UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig); +SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit, + Byte *dest, size_t *destLen, UInt32 desiredPackSize, UInt32 *unpackSize); +const Byte *LzmaEnc_GetCurBuf(CLzmaEncHandle pp); +void LzmaEnc_Finish(CLzmaEncHandle pp); +void LzmaEnc_SaveState(CLzmaEncHandle pp); +void LzmaEnc_RestoreState(CLzmaEncHandle pp); + +#ifdef SHOW_STAT +static unsigned g_STAT_OFFSET = 0; +#endif + +/* for good normalization speed we still reserve 256 MB before 4 GB range */ +#define kLzmaMaxHistorySize ((UInt32)15 << 28) + +#define kNumTopBits 24 +#define kTopValue ((UInt32)1 << kNumTopBits) + +#define kNumBitModelTotalBits 11 +#define kBitModelTotal (1 << kNumBitModelTotalBits) +#define kNumMoveBits 5 +#define kProbInitValue (kBitModelTotal >> 1) + +#define kNumMoveReducingBits 4 +#define kNumBitPriceShiftBits 4 +// #define kBitPrice (1 << kNumBitPriceShiftBits) + +#define REP_LEN_COUNT 64 + +void LzmaEncProps_Init(CLzmaEncProps *p) +{ + p->level = 5; + p->dictSize = p->mc = 0; + p->reduceSize = (UInt64)(Int64)-1; + p->lc = p->lp = p->pb = p->algo = p->fb = p->btMode = p->numHashBytes = p->numThreads = -1; + p->writeEndMark = 0; + p->affinity = 0; +} + +void LzmaEncProps_Normalize(CLzmaEncProps *p) +{ + int level = p->level; + if (level < 0) level = 5; + p->level = level; + + if (p->dictSize == 0) + p->dictSize = + ( level <= 3 ? ((UInt32)1 << (level * 2 + 16)) : + ( level <= 6 ? ((UInt32)1 << (level + 19)) : + ( level <= 7 ? ((UInt32)1 << 25) : ((UInt32)1 << 26) + ))); + + if (p->dictSize > p->reduceSize) + { + UInt32 v = (UInt32)p->reduceSize; + const UInt32 kReduceMin = ((UInt32)1 << 12); + if (v < kReduceMin) + v = kReduceMin; + if (p->dictSize > v) + p->dictSize = v; + } + + if (p->lc < 0) p->lc = 3; + if (p->lp < 0) p->lp = 0; + if (p->pb < 0) p->pb = 2; + + if (p->algo < 0) p->algo = (level < 5 ? 0 : 1); + if (p->fb < 0) p->fb = (level < 7 ? 32 : 64); + if (p->btMode < 0) p->btMode = (p->algo == 0 ? 0 : 1); + if (p->numHashBytes < 0) p->numHashBytes = (p->btMode ? 4 : 5); + if (p->mc == 0) p->mc = (16 + ((unsigned)p->fb >> 1)) >> (p->btMode ? 0 : 1); + + if (p->numThreads < 0) + p->numThreads = + #ifndef _7ZIP_ST + ((p->btMode && p->algo) ? 2 : 1); + #else + 1; + #endif +} + +UInt32 LzmaEncProps_GetDictSize(const CLzmaEncProps *props2) +{ + CLzmaEncProps props = *props2; + LzmaEncProps_Normalize(&props); + return props.dictSize; +} + + +/* +x86/x64: + +BSR: + IF (SRC == 0) ZF = 1, DEST is undefined; + AMD : DEST is unchanged; + IF (SRC != 0) ZF = 0; DEST is index of top non-zero bit + BSR is slow in some processors + +LZCNT: + IF (SRC == 0) CF = 1, DEST is size_in_bits_of_register(src) (32 or 64) + IF (SRC != 0) CF = 0, DEST = num_lead_zero_bits + IF (DEST == 0) ZF = 1; + +LZCNT works only in new processors starting from Haswell. +if LZCNT is not supported by processor, then it's executed as BSR. +LZCNT can be faster than BSR, if supported. +*/ + +// #define LZMA_LOG_BSR + +#if defined(MY_CPU_ARM_OR_ARM64) /* || defined(MY_CPU_X86_OR_AMD64) */ + + #if (defined(__clang__) && (__clang_major__ >= 6)) \ + || (defined(__GNUC__) && (__GNUC__ >= 6)) + #define LZMA_LOG_BSR + #elif defined(_MSC_VER) && (_MSC_VER >= 1300) + // #if defined(MY_CPU_ARM_OR_ARM64) + #define LZMA_LOG_BSR + // #endif + #endif +#endif + +// #include + +#ifdef LZMA_LOG_BSR + +#if defined(__clang__) \ + || defined(__GNUC__) + +/* + C code: : (30 - __builtin_clz(x)) + gcc9/gcc10 for x64 /x86 : 30 - (bsr(x) xor 31) + clang10 for x64 : 31 + (bsr(x) xor -32) +*/ + + #define MY_clz(x) ((unsigned)__builtin_clz(x)) + // __lzcnt32 + // __builtin_ia32_lzcnt_u32 + +#else // #if defined(_MSC_VER) + + #ifdef MY_CPU_ARM_OR_ARM64 + + #define MY_clz _CountLeadingZeros + + #else // if defined(MY_CPU_X86_OR_AMD64) + + // #define MY_clz __lzcnt // we can use lzcnt (unsupported by old CPU) + // _BitScanReverse code is not optimal for some MSVC compilers + #define BSR2_RET(pos, res) { unsigned long zz; _BitScanReverse(&zz, (pos)); zz--; \ + res = (zz + zz) + (pos >> zz); } + + #endif // MY_CPU_X86_OR_AMD64 + +#endif // _MSC_VER + + +#ifndef BSR2_RET + + #define BSR2_RET(pos, res) { unsigned zz = 30 - MY_clz(pos); \ + res = (zz + zz) + (pos >> zz); } + +#endif + + +unsigned GetPosSlot1(UInt32 pos); +unsigned GetPosSlot1(UInt32 pos) +{ + unsigned res; + BSR2_RET(pos, res); + return res; +} +#define GetPosSlot2(pos, res) { BSR2_RET(pos, res); } +#define GetPosSlot(pos, res) { if (pos < 2) res = pos; else BSR2_RET(pos, res); } + + +#else // ! LZMA_LOG_BSR + +#define kNumLogBits (11 + sizeof(size_t) / 8 * 3) + +#define kDicLogSizeMaxCompress ((kNumLogBits - 1) * 2 + 7) + +static void LzmaEnc_FastPosInit(Byte *g_FastPos) +{ + unsigned slot; + g_FastPos[0] = 0; + g_FastPos[1] = 1; + g_FastPos += 2; + + for (slot = 2; slot < kNumLogBits * 2; slot++) + { + size_t k = ((size_t)1 << ((slot >> 1) - 1)); + size_t j; + for (j = 0; j < k; j++) + g_FastPos[j] = (Byte)slot; + g_FastPos += k; + } +} + +/* we can use ((limit - pos) >> 31) only if (pos < ((UInt32)1 << 31)) */ +/* +#define BSR2_RET(pos, res) { unsigned zz = 6 + ((kNumLogBits - 1) & \ + (0 - (((((UInt32)1 << (kNumLogBits + 6)) - 1) - pos) >> 31))); \ + res = p->g_FastPos[pos >> zz] + (zz * 2); } +*/ + +/* +#define BSR2_RET(pos, res) { unsigned zz = 6 + ((kNumLogBits - 1) & \ + (0 - (((((UInt32)1 << (kNumLogBits)) - 1) - (pos >> 6)) >> 31))); \ + res = p->g_FastPos[pos >> zz] + (zz * 2); } +*/ + +#define BSR2_RET(pos, res) { unsigned zz = (pos < (1 << (kNumLogBits + 6))) ? 6 : 6 + kNumLogBits - 1; \ + res = p->g_FastPos[pos >> zz] + (zz * 2); } + +/* +#define BSR2_RET(pos, res) { res = (pos < (1 << (kNumLogBits + 6))) ? \ + p->g_FastPos[pos >> 6] + 12 : \ + p->g_FastPos[pos >> (6 + kNumLogBits - 1)] + (6 + (kNumLogBits - 1)) * 2; } +*/ + +#define GetPosSlot1(pos) p->g_FastPos[pos] +#define GetPosSlot2(pos, res) { BSR2_RET(pos, res); } +#define GetPosSlot(pos, res) { if (pos < kNumFullDistances) res = p->g_FastPos[pos & (kNumFullDistances - 1)]; else BSR2_RET(pos, res); } + +#endif // LZMA_LOG_BSR + + +#define LZMA_NUM_REPS 4 + +typedef UInt16 CState; +typedef UInt16 CExtra; + +typedef struct +{ + UInt32 price; + CState state; + CExtra extra; + // 0 : normal + // 1 : LIT : MATCH + // > 1 : MATCH (extra-1) : LIT : REP0 (len) + UInt32 len; + UInt32 dist; + UInt32 reps[LZMA_NUM_REPS]; +} COptimal; + + +// 18.06 +#define kNumOpts (1 << 11) +#define kPackReserve (kNumOpts * 8) +// #define kNumOpts (1 << 12) +// #define kPackReserve (1 + kNumOpts * 2) + +#define kNumLenToPosStates 4 +#define kNumPosSlotBits 6 +// #define kDicLogSizeMin 0 +#define kDicLogSizeMax 32 +#define kDistTableSizeMax (kDicLogSizeMax * 2) + +#define kNumAlignBits 4 +#define kAlignTableSize (1 << kNumAlignBits) +#define kAlignMask (kAlignTableSize - 1) + +#define kStartPosModelIndex 4 +#define kEndPosModelIndex 14 +#define kNumFullDistances (1 << (kEndPosModelIndex >> 1)) + +typedef +#ifdef _LZMA_PROB32 + UInt32 +#else + UInt16 +#endif + CLzmaProb; + +#define LZMA_PB_MAX 4 +#define LZMA_LC_MAX 8 +#define LZMA_LP_MAX 4 + +#define LZMA_NUM_PB_STATES_MAX (1 << LZMA_PB_MAX) + +#define kLenNumLowBits 3 +#define kLenNumLowSymbols (1 << kLenNumLowBits) +#define kLenNumHighBits 8 +#define kLenNumHighSymbols (1 << kLenNumHighBits) +#define kLenNumSymbolsTotal (kLenNumLowSymbols * 2 + kLenNumHighSymbols) + +#define LZMA_MATCH_LEN_MIN 2 +#define LZMA_MATCH_LEN_MAX (LZMA_MATCH_LEN_MIN + kLenNumSymbolsTotal - 1) + +#define kNumStates 12 + + +typedef struct +{ + CLzmaProb low[LZMA_NUM_PB_STATES_MAX << (kLenNumLowBits + 1)]; + CLzmaProb high[kLenNumHighSymbols]; +} CLenEnc; + + +typedef struct +{ + unsigned tableSize; + UInt32 prices[LZMA_NUM_PB_STATES_MAX][kLenNumSymbolsTotal]; + // UInt32 prices1[LZMA_NUM_PB_STATES_MAX][kLenNumLowSymbols * 2]; + // UInt32 prices2[kLenNumSymbolsTotal]; +} CLenPriceEnc; + +#define GET_PRICE_LEN(p, posState, len) \ + ((p)->prices[posState][(size_t)(len) - LZMA_MATCH_LEN_MIN]) + +/* +#define GET_PRICE_LEN(p, posState, len) \ + ((p)->prices2[(size_t)(len) - 2] + ((p)->prices1[posState][((len) - 2) & (kLenNumLowSymbols * 2 - 1)] & (((len) - 2 - kLenNumLowSymbols * 2) >> 9))) +*/ + +typedef struct +{ + UInt32 range; + unsigned cache; + UInt64 low; + UInt64 cacheSize; + Byte *buf; + Byte *bufLim; + Byte *bufBase; + ISeqOutStream *outStream; + UInt64 processed; + SRes res; +} CRangeEnc; + + +typedef struct +{ + CLzmaProb *litProbs; + + unsigned state; + UInt32 reps[LZMA_NUM_REPS]; + + CLzmaProb posAlignEncoder[1 << kNumAlignBits]; + CLzmaProb isRep[kNumStates]; + CLzmaProb isRepG0[kNumStates]; + CLzmaProb isRepG1[kNumStates]; + CLzmaProb isRepG2[kNumStates]; + CLzmaProb isMatch[kNumStates][LZMA_NUM_PB_STATES_MAX]; + CLzmaProb isRep0Long[kNumStates][LZMA_NUM_PB_STATES_MAX]; + + CLzmaProb posSlotEncoder[kNumLenToPosStates][1 << kNumPosSlotBits]; + CLzmaProb posEncoders[kNumFullDistances]; + + CLenEnc lenProbs; + CLenEnc repLenProbs; + +} CSaveState; + + +typedef UInt32 CProbPrice; + + +typedef struct +{ + void *matchFinderObj; + IMatchFinder2 matchFinder; + + unsigned optCur; + unsigned optEnd; + + unsigned longestMatchLen; + unsigned numPairs; + UInt32 numAvail; + + unsigned state; + unsigned numFastBytes; + unsigned additionalOffset; + UInt32 reps[LZMA_NUM_REPS]; + unsigned lpMask, pbMask; + CLzmaProb *litProbs; + CRangeEnc rc; + + UInt32 backRes; + + unsigned lc, lp, pb; + unsigned lclp; + + BoolInt fastMode; + BoolInt writeEndMark; + BoolInt finished; + BoolInt multiThread; + BoolInt needInit; + // BoolInt _maxMode; + + UInt64 nowPos64; + + unsigned matchPriceCount; + // unsigned alignPriceCount; + int repLenEncCounter; + + unsigned distTableSize; + + UInt32 dictSize; + SRes result; + + #ifndef _7ZIP_ST + BoolInt mtMode; + // begin of CMatchFinderMt is used in LZ thread + CMatchFinderMt matchFinderMt; + // end of CMatchFinderMt is used in BT and HASH threads + // #else + // CMatchFinder matchFinderBase; + #endif + CMatchFinder matchFinderBase; + + + // we suppose that we have 8-bytes alignment after CMatchFinder + + #ifndef _7ZIP_ST + Byte pad[128]; + #endif + + // LZ thread + CProbPrice ProbPrices[kBitModelTotal >> kNumMoveReducingBits]; + + // we want {len , dist} pairs to be 8-bytes aligned in matches array + UInt32 matches[LZMA_MATCH_LEN_MAX * 2 + 2]; + + // we want 8-bytes alignment here + UInt32 alignPrices[kAlignTableSize]; + UInt32 posSlotPrices[kNumLenToPosStates][kDistTableSizeMax]; + UInt32 distancesPrices[kNumLenToPosStates][kNumFullDistances]; + + CLzmaProb posAlignEncoder[1 << kNumAlignBits]; + CLzmaProb isRep[kNumStates]; + CLzmaProb isRepG0[kNumStates]; + CLzmaProb isRepG1[kNumStates]; + CLzmaProb isRepG2[kNumStates]; + CLzmaProb isMatch[kNumStates][LZMA_NUM_PB_STATES_MAX]; + CLzmaProb isRep0Long[kNumStates][LZMA_NUM_PB_STATES_MAX]; + CLzmaProb posSlotEncoder[kNumLenToPosStates][1 << kNumPosSlotBits]; + CLzmaProb posEncoders[kNumFullDistances]; + + CLenEnc lenProbs; + CLenEnc repLenProbs; + + #ifndef LZMA_LOG_BSR + Byte g_FastPos[1 << kNumLogBits]; + #endif + + CLenPriceEnc lenEnc; + CLenPriceEnc repLenEnc; + + COptimal opt[kNumOpts]; + + CSaveState saveState; + + // BoolInt mf_Failure; + #ifndef _7ZIP_ST + Byte pad2[128]; + #endif +} CLzmaEnc; + + +#define MFB (p->matchFinderBase) +/* +#ifndef _7ZIP_ST +#define MFB (p->matchFinderMt.MatchFinder) +#endif +*/ + +#define COPY_ARR(dest, src, arr) memcpy(dest->arr, src->arr, sizeof(src->arr)); + +void LzmaEnc_SaveState(CLzmaEncHandle pp) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + CSaveState *dest = &p->saveState; + + dest->state = p->state; + + dest->lenProbs = p->lenProbs; + dest->repLenProbs = p->repLenProbs; + + COPY_ARR(dest, p, reps); + + COPY_ARR(dest, p, posAlignEncoder); + COPY_ARR(dest, p, isRep); + COPY_ARR(dest, p, isRepG0); + COPY_ARR(dest, p, isRepG1); + COPY_ARR(dest, p, isRepG2); + COPY_ARR(dest, p, isMatch); + COPY_ARR(dest, p, isRep0Long); + COPY_ARR(dest, p, posSlotEncoder); + COPY_ARR(dest, p, posEncoders); + + memcpy(dest->litProbs, p->litProbs, ((UInt32)0x300 << p->lclp) * sizeof(CLzmaProb)); +} + + +void LzmaEnc_RestoreState(CLzmaEncHandle pp) +{ + CLzmaEnc *dest = (CLzmaEnc *)pp; + const CSaveState *p = &dest->saveState; + + dest->state = p->state; + + dest->lenProbs = p->lenProbs; + dest->repLenProbs = p->repLenProbs; + + COPY_ARR(dest, p, reps); + + COPY_ARR(dest, p, posAlignEncoder); + COPY_ARR(dest, p, isRep); + COPY_ARR(dest, p, isRepG0); + COPY_ARR(dest, p, isRepG1); + COPY_ARR(dest, p, isRepG2); + COPY_ARR(dest, p, isMatch); + COPY_ARR(dest, p, isRep0Long); + COPY_ARR(dest, p, posSlotEncoder); + COPY_ARR(dest, p, posEncoders); + + memcpy(dest->litProbs, p->litProbs, ((UInt32)0x300 << dest->lclp) * sizeof(CLzmaProb)); +} + + + +SRes LzmaEnc_SetProps(CLzmaEncHandle pp, const CLzmaEncProps *props2) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + CLzmaEncProps props = *props2; + LzmaEncProps_Normalize(&props); + + if (props.lc > LZMA_LC_MAX + || props.lp > LZMA_LP_MAX + || props.pb > LZMA_PB_MAX) + return SZ_ERROR_PARAM; + + + if (props.dictSize > kLzmaMaxHistorySize) + props.dictSize = kLzmaMaxHistorySize; + + #ifndef LZMA_LOG_BSR + { + const UInt64 dict64 = props.dictSize; + if (dict64 > ((UInt64)1 << kDicLogSizeMaxCompress)) + return SZ_ERROR_PARAM; + } + #endif + + p->dictSize = props.dictSize; + { + unsigned fb = (unsigned)props.fb; + if (fb < 5) + fb = 5; + if (fb > LZMA_MATCH_LEN_MAX) + fb = LZMA_MATCH_LEN_MAX; + p->numFastBytes = fb; + } + p->lc = (unsigned)props.lc; + p->lp = (unsigned)props.lp; + p->pb = (unsigned)props.pb; + p->fastMode = (props.algo == 0); + // p->_maxMode = True; + MFB.btMode = (Byte)(props.btMode ? 1 : 0); + { + unsigned numHashBytes = 4; + if (props.btMode) + { + if (props.numHashBytes < 2) numHashBytes = 2; + else if (props.numHashBytes < 4) numHashBytes = (unsigned)props.numHashBytes; + } + if (props.numHashBytes >= 5) numHashBytes = 5; + + MFB.numHashBytes = numHashBytes; + } + + MFB.cutValue = props.mc; + + p->writeEndMark = (BoolInt)props.writeEndMark; + + #ifndef _7ZIP_ST + /* + if (newMultiThread != _multiThread) + { + ReleaseMatchFinder(); + _multiThread = newMultiThread; + } + */ + p->multiThread = (props.numThreads > 1); + p->matchFinderMt.btSync.affinity = + p->matchFinderMt.hashSync.affinity = props.affinity; + #endif + + return SZ_OK; +} + + +void LzmaEnc_SetDataSize(CLzmaEncHandle pp, UInt64 expectedDataSiize) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + MFB.expectedDataSize = expectedDataSiize; +} + + +#define kState_Start 0 +#define kState_LitAfterMatch 4 +#define kState_LitAfterRep 5 +#define kState_MatchAfterLit 7 +#define kState_RepAfterLit 8 + +static const Byte kLiteralNextStates[kNumStates] = {0, 0, 0, 0, 1, 2, 3, 4, 5, 6, 4, 5}; +static const Byte kMatchNextStates[kNumStates] = {7, 7, 7, 7, 7, 7, 7, 10, 10, 10, 10, 10}; +static const Byte kRepNextStates[kNumStates] = {8, 8, 8, 8, 8, 8, 8, 11, 11, 11, 11, 11}; +static const Byte kShortRepNextStates[kNumStates]= {9, 9, 9, 9, 9, 9, 9, 11, 11, 11, 11, 11}; + +#define IsLitState(s) ((s) < 7) +#define GetLenToPosState2(len) (((len) < kNumLenToPosStates - 1) ? (len) : kNumLenToPosStates - 1) +#define GetLenToPosState(len) (((len) < kNumLenToPosStates + 1) ? (len) - 2 : kNumLenToPosStates - 1) + +#define kInfinityPrice (1 << 30) + +static void RangeEnc_Construct(CRangeEnc *p) +{ + p->outStream = NULL; + p->bufBase = NULL; +} + +#define RangeEnc_GetProcessed(p) ( (p)->processed + (size_t)((p)->buf - (p)->bufBase) + (p)->cacheSize) +#define RangeEnc_GetProcessed_sizet(p) ((size_t)(p)->processed + (size_t)((p)->buf - (p)->bufBase) + (size_t)(p)->cacheSize) + +#define RC_BUF_SIZE (1 << 16) + +static int RangeEnc_Alloc(CRangeEnc *p, ISzAllocPtr alloc) +{ + if (!p->bufBase) + { + p->bufBase = (Byte *)ISzAlloc_Alloc(alloc, RC_BUF_SIZE); + if (!p->bufBase) + return 0; + p->bufLim = p->bufBase + RC_BUF_SIZE; + } + return 1; +} + +static void RangeEnc_Free(CRangeEnc *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->bufBase); + p->bufBase = NULL; +} + +static void RangeEnc_Init(CRangeEnc *p) +{ + p->range = 0xFFFFFFFF; + p->cache = 0; + p->low = 0; + p->cacheSize = 0; + + p->buf = p->bufBase; + + p->processed = 0; + p->res = SZ_OK; +} + +MY_NO_INLINE static void RangeEnc_FlushStream(CRangeEnc *p) +{ + const size_t num = (size_t)(p->buf - p->bufBase); + if (p->res == SZ_OK) + { + if (num != ISeqOutStream_Write(p->outStream, p->bufBase, num)) + p->res = SZ_ERROR_WRITE; + } + p->processed += num; + p->buf = p->bufBase; +} + +MY_NO_INLINE static void MY_FAST_CALL RangeEnc_ShiftLow(CRangeEnc *p) +{ + UInt32 low = (UInt32)p->low; + unsigned high = (unsigned)(p->low >> 32); + p->low = (UInt32)(low << 8); + if (low < (UInt32)0xFF000000 || high != 0) + { + { + Byte *buf = p->buf; + *buf++ = (Byte)(p->cache + high); + p->cache = (unsigned)(low >> 24); + p->buf = buf; + if (buf == p->bufLim) + RangeEnc_FlushStream(p); + if (p->cacheSize == 0) + return; + } + high += 0xFF; + for (;;) + { + Byte *buf = p->buf; + *buf++ = (Byte)(high); + p->buf = buf; + if (buf == p->bufLim) + RangeEnc_FlushStream(p); + if (--p->cacheSize == 0) + return; + } + } + p->cacheSize++; +} + +static void RangeEnc_FlushData(CRangeEnc *p) +{ + int i; + for (i = 0; i < 5; i++) + RangeEnc_ShiftLow(p); +} + +#define RC_NORM(p) if (range < kTopValue) { range <<= 8; RangeEnc_ShiftLow(p); } + +#define RC_BIT_PRE(p, prob) \ + ttt = *(prob); \ + newBound = (range >> kNumBitModelTotalBits) * ttt; + +// #define _LZMA_ENC_USE_BRANCH + +#ifdef _LZMA_ENC_USE_BRANCH + +#define RC_BIT(p, prob, bit) { \ + RC_BIT_PRE(p, prob) \ + if (bit == 0) { range = newBound; ttt += (kBitModelTotal - ttt) >> kNumMoveBits; } \ + else { (p)->low += newBound; range -= newBound; ttt -= ttt >> kNumMoveBits; } \ + *(prob) = (CLzmaProb)ttt; \ + RC_NORM(p) \ + } + +#else + +#define RC_BIT(p, prob, bit) { \ + UInt32 mask; \ + RC_BIT_PRE(p, prob) \ + mask = 0 - (UInt32)bit; \ + range &= mask; \ + mask &= newBound; \ + range -= mask; \ + (p)->low += mask; \ + mask = (UInt32)bit - 1; \ + range += newBound & mask; \ + mask &= (kBitModelTotal - ((1 << kNumMoveBits) - 1)); \ + mask += ((1 << kNumMoveBits) - 1); \ + ttt += (UInt32)((Int32)(mask - ttt) >> kNumMoveBits); \ + *(prob) = (CLzmaProb)ttt; \ + RC_NORM(p) \ + } + +#endif + + + + +#define RC_BIT_0_BASE(p, prob) \ + range = newBound; *(prob) = (CLzmaProb)(ttt + ((kBitModelTotal - ttt) >> kNumMoveBits)); + +#define RC_BIT_1_BASE(p, prob) \ + range -= newBound; (p)->low += newBound; *(prob) = (CLzmaProb)(ttt - (ttt >> kNumMoveBits)); \ + +#define RC_BIT_0(p, prob) \ + RC_BIT_0_BASE(p, prob) \ + RC_NORM(p) + +#define RC_BIT_1(p, prob) \ + RC_BIT_1_BASE(p, prob) \ + RC_NORM(p) + +static void RangeEnc_EncodeBit_0(CRangeEnc *p, CLzmaProb *prob) +{ + UInt32 range, ttt, newBound; + range = p->range; + RC_BIT_PRE(p, prob) + RC_BIT_0(p, prob) + p->range = range; +} + +static void LitEnc_Encode(CRangeEnc *p, CLzmaProb *probs, UInt32 sym) +{ + UInt32 range = p->range; + sym |= 0x100; + do + { + UInt32 ttt, newBound; + // RangeEnc_EncodeBit(p, probs + (sym >> 8), (sym >> 7) & 1); + CLzmaProb *prob = probs + (sym >> 8); + UInt32 bit = (sym >> 7) & 1; + sym <<= 1; + RC_BIT(p, prob, bit); + } + while (sym < 0x10000); + p->range = range; +} + +static void LitEnc_EncodeMatched(CRangeEnc *p, CLzmaProb *probs, UInt32 sym, UInt32 matchByte) +{ + UInt32 range = p->range; + UInt32 offs = 0x100; + sym |= 0x100; + do + { + UInt32 ttt, newBound; + CLzmaProb *prob; + UInt32 bit; + matchByte <<= 1; + // RangeEnc_EncodeBit(p, probs + (offs + (matchByte & offs) + (sym >> 8)), (sym >> 7) & 1); + prob = probs + (offs + (matchByte & offs) + (sym >> 8)); + bit = (sym >> 7) & 1; + sym <<= 1; + offs &= ~(matchByte ^ sym); + RC_BIT(p, prob, bit); + } + while (sym < 0x10000); + p->range = range; +} + + + +static void LzmaEnc_InitPriceTables(CProbPrice *ProbPrices) +{ + UInt32 i; + for (i = 0; i < (kBitModelTotal >> kNumMoveReducingBits); i++) + { + const unsigned kCyclesBits = kNumBitPriceShiftBits; + UInt32 w = (i << kNumMoveReducingBits) + (1 << (kNumMoveReducingBits - 1)); + unsigned bitCount = 0; + unsigned j; + for (j = 0; j < kCyclesBits; j++) + { + w = w * w; + bitCount <<= 1; + while (w >= ((UInt32)1 << 16)) + { + w >>= 1; + bitCount++; + } + } + ProbPrices[i] = (CProbPrice)(((unsigned)kNumBitModelTotalBits << kCyclesBits) - 15 - bitCount); + // printf("\n%3d: %5d", i, ProbPrices[i]); + } +} + + +#define GET_PRICE(prob, bit) \ + p->ProbPrices[((prob) ^ (unsigned)(((-(int)(bit))) & (kBitModelTotal - 1))) >> kNumMoveReducingBits]; + +#define GET_PRICEa(prob, bit) \ + ProbPrices[((prob) ^ (unsigned)((-((int)(bit))) & (kBitModelTotal - 1))) >> kNumMoveReducingBits]; + +#define GET_PRICE_0(prob) p->ProbPrices[(prob) >> kNumMoveReducingBits] +#define GET_PRICE_1(prob) p->ProbPrices[((prob) ^ (kBitModelTotal - 1)) >> kNumMoveReducingBits] + +#define GET_PRICEa_0(prob) ProbPrices[(prob) >> kNumMoveReducingBits] +#define GET_PRICEa_1(prob) ProbPrices[((prob) ^ (kBitModelTotal - 1)) >> kNumMoveReducingBits] + + +static UInt32 LitEnc_GetPrice(const CLzmaProb *probs, UInt32 sym, const CProbPrice *ProbPrices) +{ + UInt32 price = 0; + sym |= 0x100; + do + { + unsigned bit = sym & 1; + sym >>= 1; + price += GET_PRICEa(probs[sym], bit); + } + while (sym >= 2); + return price; +} + + +static UInt32 LitEnc_Matched_GetPrice(const CLzmaProb *probs, UInt32 sym, UInt32 matchByte, const CProbPrice *ProbPrices) +{ + UInt32 price = 0; + UInt32 offs = 0x100; + sym |= 0x100; + do + { + matchByte <<= 1; + price += GET_PRICEa(probs[offs + (matchByte & offs) + (sym >> 8)], (sym >> 7) & 1); + sym <<= 1; + offs &= ~(matchByte ^ sym); + } + while (sym < 0x10000); + return price; +} + + +static void RcTree_ReverseEncode(CRangeEnc *rc, CLzmaProb *probs, unsigned numBits, unsigned sym) +{ + UInt32 range = rc->range; + unsigned m = 1; + do + { + UInt32 ttt, newBound; + unsigned bit = sym & 1; + // RangeEnc_EncodeBit(rc, probs + m, bit); + sym >>= 1; + RC_BIT(rc, probs + m, bit); + m = (m << 1) | bit; + } + while (--numBits); + rc->range = range; +} + + + +static void LenEnc_Init(CLenEnc *p) +{ + unsigned i; + for (i = 0; i < (LZMA_NUM_PB_STATES_MAX << (kLenNumLowBits + 1)); i++) + p->low[i] = kProbInitValue; + for (i = 0; i < kLenNumHighSymbols; i++) + p->high[i] = kProbInitValue; +} + +static void LenEnc_Encode(CLenEnc *p, CRangeEnc *rc, unsigned sym, unsigned posState) +{ + UInt32 range, ttt, newBound; + CLzmaProb *probs = p->low; + range = rc->range; + RC_BIT_PRE(rc, probs); + if (sym >= kLenNumLowSymbols) + { + RC_BIT_1(rc, probs); + probs += kLenNumLowSymbols; + RC_BIT_PRE(rc, probs); + if (sym >= kLenNumLowSymbols * 2) + { + RC_BIT_1(rc, probs); + rc->range = range; + // RcTree_Encode(rc, p->high, kLenNumHighBits, sym - kLenNumLowSymbols * 2); + LitEnc_Encode(rc, p->high, sym - kLenNumLowSymbols * 2); + return; + } + sym -= kLenNumLowSymbols; + } + + // RcTree_Encode(rc, probs + (posState << kLenNumLowBits), kLenNumLowBits, sym); + { + unsigned m; + unsigned bit; + RC_BIT_0(rc, probs); + probs += (posState << (1 + kLenNumLowBits)); + bit = (sym >> 2) ; RC_BIT(rc, probs + 1, bit); m = (1 << 1) + bit; + bit = (sym >> 1) & 1; RC_BIT(rc, probs + m, bit); m = (m << 1) + bit; + bit = sym & 1; RC_BIT(rc, probs + m, bit); + rc->range = range; + } +} + +static void SetPrices_3(const CLzmaProb *probs, UInt32 startPrice, UInt32 *prices, const CProbPrice *ProbPrices) +{ + unsigned i; + for (i = 0; i < 8; i += 2) + { + UInt32 price = startPrice; + UInt32 prob; + price += GET_PRICEa(probs[1 ], (i >> 2)); + price += GET_PRICEa(probs[2 + (i >> 2)], (i >> 1) & 1); + prob = probs[4 + (i >> 1)]; + prices[i ] = price + GET_PRICEa_0(prob); + prices[i + 1] = price + GET_PRICEa_1(prob); + } +} + + +MY_NO_INLINE static void MY_FAST_CALL LenPriceEnc_UpdateTables( + CLenPriceEnc *p, + unsigned numPosStates, + const CLenEnc *enc, + const CProbPrice *ProbPrices) +{ + UInt32 b; + + { + unsigned prob = enc->low[0]; + UInt32 a, c; + unsigned posState; + b = GET_PRICEa_1(prob); + a = GET_PRICEa_0(prob); + c = b + GET_PRICEa_0(enc->low[kLenNumLowSymbols]); + for (posState = 0; posState < numPosStates; posState++) + { + UInt32 *prices = p->prices[posState]; + const CLzmaProb *probs = enc->low + (posState << (1 + kLenNumLowBits)); + SetPrices_3(probs, a, prices, ProbPrices); + SetPrices_3(probs + kLenNumLowSymbols, c, prices + kLenNumLowSymbols, ProbPrices); + } + } + + /* + { + unsigned i; + UInt32 b; + a = GET_PRICEa_0(enc->low[0]); + for (i = 0; i < kLenNumLowSymbols; i++) + p->prices2[i] = a; + a = GET_PRICEa_1(enc->low[0]); + b = a + GET_PRICEa_0(enc->low[kLenNumLowSymbols]); + for (i = kLenNumLowSymbols; i < kLenNumLowSymbols * 2; i++) + p->prices2[i] = b; + a += GET_PRICEa_1(enc->low[kLenNumLowSymbols]); + } + */ + + // p->counter = numSymbols; + // p->counter = 64; + + { + unsigned i = p->tableSize; + + if (i > kLenNumLowSymbols * 2) + { + const CLzmaProb *probs = enc->high; + UInt32 *prices = p->prices[0] + kLenNumLowSymbols * 2; + i -= kLenNumLowSymbols * 2 - 1; + i >>= 1; + b += GET_PRICEa_1(enc->low[kLenNumLowSymbols]); + do + { + /* + p->prices2[i] = a + + // RcTree_GetPrice(enc->high, kLenNumHighBits, i - kLenNumLowSymbols * 2, ProbPrices); + LitEnc_GetPrice(probs, i - kLenNumLowSymbols * 2, ProbPrices); + */ + // UInt32 price = a + RcTree_GetPrice(probs, kLenNumHighBits - 1, sym, ProbPrices); + unsigned sym = --i + (1 << (kLenNumHighBits - 1)); + UInt32 price = b; + do + { + unsigned bit = sym & 1; + sym >>= 1; + price += GET_PRICEa(probs[sym], bit); + } + while (sym >= 2); + + { + unsigned prob = probs[(size_t)i + (1 << (kLenNumHighBits - 1))]; + prices[(size_t)i * 2 ] = price + GET_PRICEa_0(prob); + prices[(size_t)i * 2 + 1] = price + GET_PRICEa_1(prob); + } + } + while (i); + + { + unsigned posState; + size_t num = (p->tableSize - kLenNumLowSymbols * 2) * sizeof(p->prices[0][0]); + for (posState = 1; posState < numPosStates; posState++) + memcpy(p->prices[posState] + kLenNumLowSymbols * 2, p->prices[0] + kLenNumLowSymbols * 2, num); + } + } + } +} + +/* + #ifdef SHOW_STAT + g_STAT_OFFSET += num; + printf("\n MovePos %u", num); + #endif +*/ + +#define MOVE_POS(p, num) { \ + p->additionalOffset += (num); \ + p->matchFinder.Skip(p->matchFinderObj, (UInt32)(num)); } + + +static unsigned ReadMatchDistances(CLzmaEnc *p, unsigned *numPairsRes) +{ + unsigned numPairs; + + p->additionalOffset++; + p->numAvail = p->matchFinder.GetNumAvailableBytes(p->matchFinderObj); + { + const UInt32 *d = p->matchFinder.GetMatches(p->matchFinderObj, p->matches); + // if (!d) { p->mf_Failure = True; *numPairsRes = 0; return 0; } + numPairs = (unsigned)(d - p->matches); + } + *numPairsRes = numPairs; + + #ifdef SHOW_STAT + printf("\n i = %u numPairs = %u ", g_STAT_OFFSET, numPairs / 2); + g_STAT_OFFSET++; + { + unsigned i; + for (i = 0; i < numPairs; i += 2) + printf("%2u %6u | ", p->matches[i], p->matches[i + 1]); + } + #endif + + if (numPairs == 0) + return 0; + { + const unsigned len = p->matches[(size_t)numPairs - 2]; + if (len != p->numFastBytes) + return len; + { + UInt32 numAvail = p->numAvail; + if (numAvail > LZMA_MATCH_LEN_MAX) + numAvail = LZMA_MATCH_LEN_MAX; + { + const Byte *p1 = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + const Byte *p2 = p1 + len; + const ptrdiff_t dif = (ptrdiff_t)-1 - (ptrdiff_t)p->matches[(size_t)numPairs - 1]; + const Byte *lim = p1 + numAvail; + for (; p2 != lim && *p2 == p2[dif]; p2++) + {} + return (unsigned)(p2 - p1); + } + } + } +} + +#define MARK_LIT ((UInt32)(Int32)-1) + +#define MakeAs_Lit(p) { (p)->dist = MARK_LIT; (p)->extra = 0; } +#define MakeAs_ShortRep(p) { (p)->dist = 0; (p)->extra = 0; } +#define IsShortRep(p) ((p)->dist == 0) + + +#define GetPrice_ShortRep(p, state, posState) \ + ( GET_PRICE_0(p->isRepG0[state]) + GET_PRICE_0(p->isRep0Long[state][posState])) + +#define GetPrice_Rep_0(p, state, posState) ( \ + GET_PRICE_1(p->isMatch[state][posState]) \ + + GET_PRICE_1(p->isRep0Long[state][posState])) \ + + GET_PRICE_1(p->isRep[state]) \ + + GET_PRICE_0(p->isRepG0[state]) + +MY_FORCE_INLINE +static UInt32 GetPrice_PureRep(const CLzmaEnc *p, unsigned repIndex, size_t state, size_t posState) +{ + UInt32 price; + UInt32 prob = p->isRepG0[state]; + if (repIndex == 0) + { + price = GET_PRICE_0(prob); + price += GET_PRICE_1(p->isRep0Long[state][posState]); + } + else + { + price = GET_PRICE_1(prob); + prob = p->isRepG1[state]; + if (repIndex == 1) + price += GET_PRICE_0(prob); + else + { + price += GET_PRICE_1(prob); + price += GET_PRICE(p->isRepG2[state], repIndex - 2); + } + } + return price; +} + + +static unsigned Backward(CLzmaEnc *p, unsigned cur) +{ + unsigned wr = cur + 1; + p->optEnd = wr; + + for (;;) + { + UInt32 dist = p->opt[cur].dist; + unsigned len = (unsigned)p->opt[cur].len; + unsigned extra = (unsigned)p->opt[cur].extra; + cur -= len; + + if (extra) + { + wr--; + p->opt[wr].len = (UInt32)len; + cur -= extra; + len = extra; + if (extra == 1) + { + p->opt[wr].dist = dist; + dist = MARK_LIT; + } + else + { + p->opt[wr].dist = 0; + len--; + wr--; + p->opt[wr].dist = MARK_LIT; + p->opt[wr].len = 1; + } + } + + if (cur == 0) + { + p->backRes = dist; + p->optCur = wr; + return len; + } + + wr--; + p->opt[wr].dist = dist; + p->opt[wr].len = (UInt32)len; + } +} + + + +#define LIT_PROBS(pos, prevByte) \ + (p->litProbs + (UInt32)3 * (((((pos) << 8) + (prevByte)) & p->lpMask) << p->lc)) + + +static unsigned GetOptimum(CLzmaEnc *p, UInt32 position) +{ + unsigned last, cur; + UInt32 reps[LZMA_NUM_REPS]; + unsigned repLens[LZMA_NUM_REPS]; + UInt32 *matches; + + { + UInt32 numAvail; + unsigned numPairs, mainLen, repMaxIndex, i, posState; + UInt32 matchPrice, repMatchPrice; + const Byte *data; + Byte curByte, matchByte; + + p->optCur = p->optEnd = 0; + + if (p->additionalOffset == 0) + mainLen = ReadMatchDistances(p, &numPairs); + else + { + mainLen = p->longestMatchLen; + numPairs = p->numPairs; + } + + numAvail = p->numAvail; + if (numAvail < 2) + { + p->backRes = MARK_LIT; + return 1; + } + if (numAvail > LZMA_MATCH_LEN_MAX) + numAvail = LZMA_MATCH_LEN_MAX; + + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + repMaxIndex = 0; + + for (i = 0; i < LZMA_NUM_REPS; i++) + { + unsigned len; + const Byte *data2; + reps[i] = p->reps[i]; + data2 = data - reps[i]; + if (data[0] != data2[0] || data[1] != data2[1]) + { + repLens[i] = 0; + continue; + } + for (len = 2; len < numAvail && data[len] == data2[len]; len++) + {} + repLens[i] = len; + if (len > repLens[repMaxIndex]) + repMaxIndex = i; + if (len == LZMA_MATCH_LEN_MAX) // 21.03 : optimization + break; + } + + if (repLens[repMaxIndex] >= p->numFastBytes) + { + unsigned len; + p->backRes = (UInt32)repMaxIndex; + len = repLens[repMaxIndex]; + MOVE_POS(p, len - 1) + return len; + } + + matches = p->matches; + #define MATCHES matches + // #define MATCHES p->matches + + if (mainLen >= p->numFastBytes) + { + p->backRes = MATCHES[(size_t)numPairs - 1] + LZMA_NUM_REPS; + MOVE_POS(p, mainLen - 1) + return mainLen; + } + + curByte = *data; + matchByte = *(data - reps[0]); + + last = repLens[repMaxIndex]; + if (last <= mainLen) + last = mainLen; + + if (last < 2 && curByte != matchByte) + { + p->backRes = MARK_LIT; + return 1; + } + + p->opt[0].state = (CState)p->state; + + posState = (position & p->pbMask); + + { + const CLzmaProb *probs = LIT_PROBS(position, *(data - 1)); + p->opt[1].price = GET_PRICE_0(p->isMatch[p->state][posState]) + + (!IsLitState(p->state) ? + LitEnc_Matched_GetPrice(probs, curByte, matchByte, p->ProbPrices) : + LitEnc_GetPrice(probs, curByte, p->ProbPrices)); + } + + MakeAs_Lit(&p->opt[1]); + + matchPrice = GET_PRICE_1(p->isMatch[p->state][posState]); + repMatchPrice = matchPrice + GET_PRICE_1(p->isRep[p->state]); + + // 18.06 + if (matchByte == curByte && repLens[0] == 0) + { + UInt32 shortRepPrice = repMatchPrice + GetPrice_ShortRep(p, p->state, posState); + if (shortRepPrice < p->opt[1].price) + { + p->opt[1].price = shortRepPrice; + MakeAs_ShortRep(&p->opt[1]); + } + if (last < 2) + { + p->backRes = p->opt[1].dist; + return 1; + } + } + + p->opt[1].len = 1; + + p->opt[0].reps[0] = reps[0]; + p->opt[0].reps[1] = reps[1]; + p->opt[0].reps[2] = reps[2]; + p->opt[0].reps[3] = reps[3]; + + // ---------- REP ---------- + + for (i = 0; i < LZMA_NUM_REPS; i++) + { + unsigned repLen = repLens[i]; + UInt32 price; + if (repLen < 2) + continue; + price = repMatchPrice + GetPrice_PureRep(p, i, p->state, posState); + do + { + UInt32 price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState, repLen); + COptimal *opt = &p->opt[repLen]; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)repLen; + opt->dist = (UInt32)i; + opt->extra = 0; + } + } + while (--repLen >= 2); + } + + + // ---------- MATCH ---------- + { + unsigned len = repLens[0] + 1; + if (len <= mainLen) + { + unsigned offs = 0; + UInt32 normalMatchPrice = matchPrice + GET_PRICE_0(p->isRep[p->state]); + + if (len < 2) + len = 2; + else + while (len > MATCHES[offs]) + offs += 2; + + for (; ; len++) + { + COptimal *opt; + UInt32 dist = MATCHES[(size_t)offs + 1]; + UInt32 price = normalMatchPrice + GET_PRICE_LEN(&p->lenEnc, posState, len); + unsigned lenToPosState = GetLenToPosState(len); + + if (dist < kNumFullDistances) + price += p->distancesPrices[lenToPosState][dist & (kNumFullDistances - 1)]; + else + { + unsigned slot; + GetPosSlot2(dist, slot); + price += p->alignPrices[dist & kAlignMask]; + price += p->posSlotPrices[lenToPosState][slot]; + } + + opt = &p->opt[len]; + + if (price < opt->price) + { + opt->price = price; + opt->len = (UInt32)len; + opt->dist = dist + LZMA_NUM_REPS; + opt->extra = 0; + } + + if (len == MATCHES[offs]) + { + offs += 2; + if (offs == numPairs) + break; + } + } + } + } + + + cur = 0; + + #ifdef SHOW_STAT2 + /* if (position >= 0) */ + { + unsigned i; + printf("\n pos = %4X", position); + for (i = cur; i <= last; i++) + printf("\nprice[%4X] = %u", position - cur + i, p->opt[i].price); + } + #endif + } + + + + // ---------- Optimal Parsing ---------- + + for (;;) + { + unsigned numAvail; + UInt32 numAvailFull; + unsigned newLen, numPairs, prev, state, posState, startLen; + UInt32 litPrice, matchPrice, repMatchPrice; + BoolInt nextIsLit; + Byte curByte, matchByte; + const Byte *data; + COptimal *curOpt, *nextOpt; + + if (++cur == last) + break; + + // 18.06 + if (cur >= kNumOpts - 64) + { + unsigned j, best; + UInt32 price = p->opt[cur].price; + best = cur; + for (j = cur + 1; j <= last; j++) + { + UInt32 price2 = p->opt[j].price; + if (price >= price2) + { + price = price2; + best = j; + } + } + { + unsigned delta = best - cur; + if (delta != 0) + { + MOVE_POS(p, delta); + } + } + cur = best; + break; + } + + newLen = ReadMatchDistances(p, &numPairs); + + if (newLen >= p->numFastBytes) + { + p->numPairs = numPairs; + p->longestMatchLen = newLen; + break; + } + + curOpt = &p->opt[cur]; + + position++; + + // we need that check here, if skip_items in p->opt are possible + /* + if (curOpt->price >= kInfinityPrice) + continue; + */ + + prev = cur - curOpt->len; + + if (curOpt->len == 1) + { + state = (unsigned)p->opt[prev].state; + if (IsShortRep(curOpt)) + state = kShortRepNextStates[state]; + else + state = kLiteralNextStates[state]; + } + else + { + const COptimal *prevOpt; + UInt32 b0; + UInt32 dist = curOpt->dist; + + if (curOpt->extra) + { + prev -= (unsigned)curOpt->extra; + state = kState_RepAfterLit; + if (curOpt->extra == 1) + state = (dist < LZMA_NUM_REPS ? kState_RepAfterLit : kState_MatchAfterLit); + } + else + { + state = (unsigned)p->opt[prev].state; + if (dist < LZMA_NUM_REPS) + state = kRepNextStates[state]; + else + state = kMatchNextStates[state]; + } + + prevOpt = &p->opt[prev]; + b0 = prevOpt->reps[0]; + + if (dist < LZMA_NUM_REPS) + { + if (dist == 0) + { + reps[0] = b0; + reps[1] = prevOpt->reps[1]; + reps[2] = prevOpt->reps[2]; + reps[3] = prevOpt->reps[3]; + } + else + { + reps[1] = b0; + b0 = prevOpt->reps[1]; + if (dist == 1) + { + reps[0] = b0; + reps[2] = prevOpt->reps[2]; + reps[3] = prevOpt->reps[3]; + } + else + { + reps[2] = b0; + reps[0] = prevOpt->reps[dist]; + reps[3] = prevOpt->reps[dist ^ 1]; + } + } + } + else + { + reps[0] = (dist - LZMA_NUM_REPS + 1); + reps[1] = b0; + reps[2] = prevOpt->reps[1]; + reps[3] = prevOpt->reps[2]; + } + } + + curOpt->state = (CState)state; + curOpt->reps[0] = reps[0]; + curOpt->reps[1] = reps[1]; + curOpt->reps[2] = reps[2]; + curOpt->reps[3] = reps[3]; + + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + curByte = *data; + matchByte = *(data - reps[0]); + + posState = (position & p->pbMask); + + /* + The order of Price checks: + < LIT + <= SHORT_REP + < LIT : REP_0 + < REP [ : LIT : REP_0 ] + < MATCH [ : LIT : REP_0 ] + */ + + { + UInt32 curPrice = curOpt->price; + unsigned prob = p->isMatch[state][posState]; + matchPrice = curPrice + GET_PRICE_1(prob); + litPrice = curPrice + GET_PRICE_0(prob); + } + + nextOpt = &p->opt[(size_t)cur + 1]; + nextIsLit = False; + + // here we can allow skip_items in p->opt, if we don't check (nextOpt->price < kInfinityPrice) + // 18.new.06 + if ((nextOpt->price < kInfinityPrice + // && !IsLitState(state) + && matchByte == curByte) + || litPrice > nextOpt->price + ) + litPrice = 0; + else + { + const CLzmaProb *probs = LIT_PROBS(position, *(data - 1)); + litPrice += (!IsLitState(state) ? + LitEnc_Matched_GetPrice(probs, curByte, matchByte, p->ProbPrices) : + LitEnc_GetPrice(probs, curByte, p->ProbPrices)); + + if (litPrice < nextOpt->price) + { + nextOpt->price = litPrice; + nextOpt->len = 1; + MakeAs_Lit(nextOpt); + nextIsLit = True; + } + } + + repMatchPrice = matchPrice + GET_PRICE_1(p->isRep[state]); + + numAvailFull = p->numAvail; + { + unsigned temp = kNumOpts - 1 - cur; + if (numAvailFull > temp) + numAvailFull = (UInt32)temp; + } + + // 18.06 + // ---------- SHORT_REP ---------- + if (IsLitState(state)) // 18.new + if (matchByte == curByte) + if (repMatchPrice < nextOpt->price) // 18.new + // if (numAvailFull < 2 || data[1] != *(data - reps[0] + 1)) + if ( + // nextOpt->price >= kInfinityPrice || + nextOpt->len < 2 // we can check nextOpt->len, if skip items are not allowed in p->opt + || (nextOpt->dist != 0 + // && nextOpt->extra <= 1 // 17.old + ) + ) + { + UInt32 shortRepPrice = repMatchPrice + GetPrice_ShortRep(p, state, posState); + // if (shortRepPrice <= nextOpt->price) // 17.old + if (shortRepPrice < nextOpt->price) // 18.new + { + nextOpt->price = shortRepPrice; + nextOpt->len = 1; + MakeAs_ShortRep(nextOpt); + nextIsLit = False; + } + } + + if (numAvailFull < 2) + continue; + numAvail = (numAvailFull <= p->numFastBytes ? numAvailFull : p->numFastBytes); + + // numAvail <= p->numFastBytes + + // ---------- LIT : REP_0 ---------- + + if (!nextIsLit + && litPrice != 0 // 18.new + && matchByte != curByte + && numAvailFull > 2) + { + const Byte *data2 = data - reps[0]; + if (data[1] == data2[1] && data[2] == data2[2]) + { + unsigned len; + unsigned limit = p->numFastBytes + 1; + if (limit > numAvailFull) + limit = numAvailFull; + for (len = 3; len < limit && data[len] == data2[len]; len++) + {} + + { + unsigned state2 = kLiteralNextStates[state]; + unsigned posState2 = (position + 1) & p->pbMask; + UInt32 price = litPrice + GetPrice_Rep_0(p, state2, posState2); + { + unsigned offset = cur + len; + + if (last < offset) + last = offset; + + // do + { + UInt32 price2; + COptimal *opt; + len--; + // price2 = price + GetPrice_Len_Rep_0(p, len, state2, posState2); + price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState2, len); + + opt = &p->opt[offset]; + // offset--; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)len; + opt->dist = 0; + opt->extra = 1; + } + } + // while (len >= 3); + } + } + } + } + + startLen = 2; /* speed optimization */ + + { + // ---------- REP ---------- + unsigned repIndex = 0; // 17.old + // unsigned repIndex = IsLitState(state) ? 0 : 1; // 18.notused + for (; repIndex < LZMA_NUM_REPS; repIndex++) + { + unsigned len; + UInt32 price; + const Byte *data2 = data - reps[repIndex]; + if (data[0] != data2[0] || data[1] != data2[1]) + continue; + + for (len = 2; len < numAvail && data[len] == data2[len]; len++) + {} + + // if (len < startLen) continue; // 18.new: speed optimization + + { + unsigned offset = cur + len; + if (last < offset) + last = offset; + } + { + unsigned len2 = len; + price = repMatchPrice + GetPrice_PureRep(p, repIndex, state, posState); + do + { + UInt32 price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState, len2); + COptimal *opt = &p->opt[cur + len2]; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)len2; + opt->dist = (UInt32)repIndex; + opt->extra = 0; + } + } + while (--len2 >= 2); + } + + if (repIndex == 0) startLen = len + 1; // 17.old + // startLen = len + 1; // 18.new + + /* if (_maxMode) */ + { + // ---------- REP : LIT : REP_0 ---------- + // numFastBytes + 1 + numFastBytes + + unsigned len2 = len + 1; + unsigned limit = len2 + p->numFastBytes; + if (limit > numAvailFull) + limit = numAvailFull; + + len2 += 2; + if (len2 <= limit) + if (data[len2 - 2] == data2[len2 - 2]) + if (data[len2 - 1] == data2[len2 - 1]) + { + unsigned state2 = kRepNextStates[state]; + unsigned posState2 = (position + len) & p->pbMask; + price += GET_PRICE_LEN(&p->repLenEnc, posState, len) + + GET_PRICE_0(p->isMatch[state2][posState2]) + + LitEnc_Matched_GetPrice(LIT_PROBS(position + len, data[(size_t)len - 1]), + data[len], data2[len], p->ProbPrices); + + // state2 = kLiteralNextStates[state2]; + state2 = kState_LitAfterRep; + posState2 = (posState2 + 1) & p->pbMask; + + + price += GetPrice_Rep_0(p, state2, posState2); + + for (; len2 < limit && data[len2] == data2[len2]; len2++) + {} + + len2 -= len; + // if (len2 >= 3) + { + { + unsigned offset = cur + len + len2; + + if (last < offset) + last = offset; + // do + { + UInt32 price2; + COptimal *opt; + len2--; + // price2 = price + GetPrice_Len_Rep_0(p, len2, state2, posState2); + price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState2, len2); + + opt = &p->opt[offset]; + // offset--; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)len2; + opt->extra = (CExtra)(len + 1); + opt->dist = (UInt32)repIndex; + } + } + // while (len2 >= 3); + } + } + } + } + } + } + + + // ---------- MATCH ---------- + /* for (unsigned len = 2; len <= newLen; len++) */ + if (newLen > numAvail) + { + newLen = numAvail; + for (numPairs = 0; newLen > MATCHES[numPairs]; numPairs += 2); + MATCHES[numPairs] = (UInt32)newLen; + numPairs += 2; + } + + // startLen = 2; /* speed optimization */ + + if (newLen >= startLen) + { + UInt32 normalMatchPrice = matchPrice + GET_PRICE_0(p->isRep[state]); + UInt32 dist; + unsigned offs, posSlot, len; + + { + unsigned offset = cur + newLen; + if (last < offset) + last = offset; + } + + offs = 0; + while (startLen > MATCHES[offs]) + offs += 2; + dist = MATCHES[(size_t)offs + 1]; + + // if (dist >= kNumFullDistances) + GetPosSlot2(dist, posSlot); + + for (len = /*2*/ startLen; ; len++) + { + UInt32 price = normalMatchPrice + GET_PRICE_LEN(&p->lenEnc, posState, len); + { + COptimal *opt; + unsigned lenNorm = len - 2; + lenNorm = GetLenToPosState2(lenNorm); + if (dist < kNumFullDistances) + price += p->distancesPrices[lenNorm][dist & (kNumFullDistances - 1)]; + else + price += p->posSlotPrices[lenNorm][posSlot] + p->alignPrices[dist & kAlignMask]; + + opt = &p->opt[cur + len]; + if (price < opt->price) + { + opt->price = price; + opt->len = (UInt32)len; + opt->dist = dist + LZMA_NUM_REPS; + opt->extra = 0; + } + } + + if (len == MATCHES[offs]) + { + // if (p->_maxMode) { + // MATCH : LIT : REP_0 + + const Byte *data2 = data - dist - 1; + unsigned len2 = len + 1; + unsigned limit = len2 + p->numFastBytes; + if (limit > numAvailFull) + limit = numAvailFull; + + len2 += 2; + if (len2 <= limit) + if (data[len2 - 2] == data2[len2 - 2]) + if (data[len2 - 1] == data2[len2 - 1]) + { + for (; len2 < limit && data[len2] == data2[len2]; len2++) + {} + + len2 -= len; + + // if (len2 >= 3) + { + unsigned state2 = kMatchNextStates[state]; + unsigned posState2 = (position + len) & p->pbMask; + unsigned offset; + price += GET_PRICE_0(p->isMatch[state2][posState2]); + price += LitEnc_Matched_GetPrice(LIT_PROBS(position + len, data[(size_t)len - 1]), + data[len], data2[len], p->ProbPrices); + + // state2 = kLiteralNextStates[state2]; + state2 = kState_LitAfterMatch; + + posState2 = (posState2 + 1) & p->pbMask; + price += GetPrice_Rep_0(p, state2, posState2); + + offset = cur + len + len2; + + if (last < offset) + last = offset; + // do + { + UInt32 price2; + COptimal *opt; + len2--; + // price2 = price + GetPrice_Len_Rep_0(p, len2, state2, posState2); + price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState2, len2); + opt = &p->opt[offset]; + // offset--; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)len2; + opt->extra = (CExtra)(len + 1); + opt->dist = dist + LZMA_NUM_REPS; + } + } + // while (len2 >= 3); + } + + } + + offs += 2; + if (offs == numPairs) + break; + dist = MATCHES[(size_t)offs + 1]; + // if (dist >= kNumFullDistances) + GetPosSlot2(dist, posSlot); + } + } + } + } + + do + p->opt[last].price = kInfinityPrice; + while (--last); + + return Backward(p, cur); +} + + + +#define ChangePair(smallDist, bigDist) (((bigDist) >> 7) > (smallDist)) + + + +static unsigned GetOptimumFast(CLzmaEnc *p) +{ + UInt32 numAvail, mainDist; + unsigned mainLen, numPairs, repIndex, repLen, i; + const Byte *data; + + if (p->additionalOffset == 0) + mainLen = ReadMatchDistances(p, &numPairs); + else + { + mainLen = p->longestMatchLen; + numPairs = p->numPairs; + } + + numAvail = p->numAvail; + p->backRes = MARK_LIT; + if (numAvail < 2) + return 1; + // if (mainLen < 2 && p->state == 0) return 1; // 18.06.notused + if (numAvail > LZMA_MATCH_LEN_MAX) + numAvail = LZMA_MATCH_LEN_MAX; + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + repLen = repIndex = 0; + + for (i = 0; i < LZMA_NUM_REPS; i++) + { + unsigned len; + const Byte *data2 = data - p->reps[i]; + if (data[0] != data2[0] || data[1] != data2[1]) + continue; + for (len = 2; len < numAvail && data[len] == data2[len]; len++) + {} + if (len >= p->numFastBytes) + { + p->backRes = (UInt32)i; + MOVE_POS(p, len - 1) + return len; + } + if (len > repLen) + { + repIndex = i; + repLen = len; + } + } + + if (mainLen >= p->numFastBytes) + { + p->backRes = p->matches[(size_t)numPairs - 1] + LZMA_NUM_REPS; + MOVE_POS(p, mainLen - 1) + return mainLen; + } + + mainDist = 0; /* for GCC */ + + if (mainLen >= 2) + { + mainDist = p->matches[(size_t)numPairs - 1]; + while (numPairs > 2) + { + UInt32 dist2; + if (mainLen != p->matches[(size_t)numPairs - 4] + 1) + break; + dist2 = p->matches[(size_t)numPairs - 3]; + if (!ChangePair(dist2, mainDist)) + break; + numPairs -= 2; + mainLen--; + mainDist = dist2; + } + if (mainLen == 2 && mainDist >= 0x80) + mainLen = 1; + } + + if (repLen >= 2) + if ( repLen + 1 >= mainLen + || (repLen + 2 >= mainLen && mainDist >= (1 << 9)) + || (repLen + 3 >= mainLen && mainDist >= (1 << 15))) + { + p->backRes = (UInt32)repIndex; + MOVE_POS(p, repLen - 1) + return repLen; + } + + if (mainLen < 2 || numAvail <= 2) + return 1; + + { + unsigned len1 = ReadMatchDistances(p, &p->numPairs); + p->longestMatchLen = len1; + + if (len1 >= 2) + { + UInt32 newDist = p->matches[(size_t)p->numPairs - 1]; + if ( (len1 >= mainLen && newDist < mainDist) + || (len1 == mainLen + 1 && !ChangePair(mainDist, newDist)) + || (len1 > mainLen + 1) + || (len1 + 1 >= mainLen && mainLen >= 3 && ChangePair(newDist, mainDist))) + return 1; + } + } + + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + + for (i = 0; i < LZMA_NUM_REPS; i++) + { + unsigned len, limit; + const Byte *data2 = data - p->reps[i]; + if (data[0] != data2[0] || data[1] != data2[1]) + continue; + limit = mainLen - 1; + for (len = 2;; len++) + { + if (len >= limit) + return 1; + if (data[len] != data2[len]) + break; + } + } + + p->backRes = mainDist + LZMA_NUM_REPS; + if (mainLen != 2) + { + MOVE_POS(p, mainLen - 2) + } + return mainLen; +} + + + + +static void WriteEndMarker(CLzmaEnc *p, unsigned posState) +{ + UInt32 range; + range = p->rc.range; + { + UInt32 ttt, newBound; + CLzmaProb *prob = &p->isMatch[p->state][posState]; + RC_BIT_PRE(&p->rc, prob) + RC_BIT_1(&p->rc, prob) + prob = &p->isRep[p->state]; + RC_BIT_PRE(&p->rc, prob) + RC_BIT_0(&p->rc, prob) + } + p->state = kMatchNextStates[p->state]; + + p->rc.range = range; + LenEnc_Encode(&p->lenProbs, &p->rc, 0, posState); + range = p->rc.range; + + { + // RcTree_Encode_PosSlot(&p->rc, p->posSlotEncoder[0], (1 << kNumPosSlotBits) - 1); + CLzmaProb *probs = p->posSlotEncoder[0]; + unsigned m = 1; + do + { + UInt32 ttt, newBound; + RC_BIT_PRE(p, probs + m) + RC_BIT_1(&p->rc, probs + m); + m = (m << 1) + 1; + } + while (m < (1 << kNumPosSlotBits)); + } + { + // RangeEnc_EncodeDirectBits(&p->rc, ((UInt32)1 << (30 - kNumAlignBits)) - 1, 30 - kNumAlignBits); UInt32 range = p->range; + unsigned numBits = 30 - kNumAlignBits; + do + { + range >>= 1; + p->rc.low += range; + RC_NORM(&p->rc) + } + while (--numBits); + } + + { + // RcTree_ReverseEncode(&p->rc, p->posAlignEncoder, kNumAlignBits, kAlignMask); + CLzmaProb *probs = p->posAlignEncoder; + unsigned m = 1; + do + { + UInt32 ttt, newBound; + RC_BIT_PRE(p, probs + m) + RC_BIT_1(&p->rc, probs + m); + m = (m << 1) + 1; + } + while (m < kAlignTableSize); + } + p->rc.range = range; +} + + +static SRes CheckErrors(CLzmaEnc *p) +{ + if (p->result != SZ_OK) + return p->result; + if (p->rc.res != SZ_OK) + p->result = SZ_ERROR_WRITE; + + #ifndef _7ZIP_ST + if ( + // p->mf_Failure || + (p->mtMode && + ( // p->matchFinderMt.failure_LZ_LZ || + p->matchFinderMt.failure_LZ_BT)) + ) + { + p->result = MY_HRES_ERROR__INTERNAL_ERROR; + // printf("\nCheckErrors p->matchFinderMt.failureLZ\n"); + } + #endif + + if (MFB.result != SZ_OK) + p->result = SZ_ERROR_READ; + + if (p->result != SZ_OK) + p->finished = True; + return p->result; +} + + +MY_NO_INLINE static SRes Flush(CLzmaEnc *p, UInt32 nowPos) +{ + /* ReleaseMFStream(); */ + p->finished = True; + if (p->writeEndMark) + WriteEndMarker(p, nowPos & p->pbMask); + RangeEnc_FlushData(&p->rc); + RangeEnc_FlushStream(&p->rc); + return CheckErrors(p); +} + + +MY_NO_INLINE static void FillAlignPrices(CLzmaEnc *p) +{ + unsigned i; + const CProbPrice *ProbPrices = p->ProbPrices; + const CLzmaProb *probs = p->posAlignEncoder; + // p->alignPriceCount = 0; + for (i = 0; i < kAlignTableSize / 2; i++) + { + UInt32 price = 0; + unsigned sym = i; + unsigned m = 1; + unsigned bit; + UInt32 prob; + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[m], bit); m = (m << 1) + bit; + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[m], bit); m = (m << 1) + bit; + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[m], bit); m = (m << 1) + bit; + prob = probs[m]; + p->alignPrices[i ] = price + GET_PRICEa_0(prob); + p->alignPrices[i + 8] = price + GET_PRICEa_1(prob); + // p->alignPrices[i] = RcTree_ReverseGetPrice(p->posAlignEncoder, kNumAlignBits, i, p->ProbPrices); + } +} + + +MY_NO_INLINE static void FillDistancesPrices(CLzmaEnc *p) +{ + // int y; for (y = 0; y < 100; y++) { + + UInt32 tempPrices[kNumFullDistances]; + unsigned i, lps; + + const CProbPrice *ProbPrices = p->ProbPrices; + p->matchPriceCount = 0; + + for (i = kStartPosModelIndex / 2; i < kNumFullDistances / 2; i++) + { + unsigned posSlot = GetPosSlot1(i); + unsigned footerBits = (posSlot >> 1) - 1; + unsigned base = ((2 | (posSlot & 1)) << footerBits); + const CLzmaProb *probs = p->posEncoders + (size_t)base * 2; + // tempPrices[i] = RcTree_ReverseGetPrice(p->posEncoders + base, footerBits, i - base, p->ProbPrices); + UInt32 price = 0; + unsigned m = 1; + unsigned sym = i; + unsigned offset = (unsigned)1 << footerBits; + base += i; + + if (footerBits) + do + { + unsigned bit = sym & 1; + sym >>= 1; + price += GET_PRICEa(probs[m], bit); + m = (m << 1) + bit; + } + while (--footerBits); + + { + unsigned prob = probs[m]; + tempPrices[base ] = price + GET_PRICEa_0(prob); + tempPrices[base + offset] = price + GET_PRICEa_1(prob); + } + } + + for (lps = 0; lps < kNumLenToPosStates; lps++) + { + unsigned slot; + unsigned distTableSize2 = (p->distTableSize + 1) >> 1; + UInt32 *posSlotPrices = p->posSlotPrices[lps]; + const CLzmaProb *probs = p->posSlotEncoder[lps]; + + for (slot = 0; slot < distTableSize2; slot++) + { + // posSlotPrices[slot] = RcTree_GetPrice(encoder, kNumPosSlotBits, slot, p->ProbPrices); + UInt32 price; + unsigned bit; + unsigned sym = slot + (1 << (kNumPosSlotBits - 1)); + unsigned prob; + bit = sym & 1; sym >>= 1; price = GET_PRICEa(probs[sym], bit); + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); + prob = probs[(size_t)slot + (1 << (kNumPosSlotBits - 1))]; + posSlotPrices[(size_t)slot * 2 ] = price + GET_PRICEa_0(prob); + posSlotPrices[(size_t)slot * 2 + 1] = price + GET_PRICEa_1(prob); + } + + { + UInt32 delta = ((UInt32)((kEndPosModelIndex / 2 - 1) - kNumAlignBits) << kNumBitPriceShiftBits); + for (slot = kEndPosModelIndex / 2; slot < distTableSize2; slot++) + { + posSlotPrices[(size_t)slot * 2 ] += delta; + posSlotPrices[(size_t)slot * 2 + 1] += delta; + delta += ((UInt32)1 << kNumBitPriceShiftBits); + } + } + + { + UInt32 *dp = p->distancesPrices[lps]; + + dp[0] = posSlotPrices[0]; + dp[1] = posSlotPrices[1]; + dp[2] = posSlotPrices[2]; + dp[3] = posSlotPrices[3]; + + for (i = 4; i < kNumFullDistances; i += 2) + { + UInt32 slotPrice = posSlotPrices[GetPosSlot1(i)]; + dp[i ] = slotPrice + tempPrices[i]; + dp[i + 1] = slotPrice + tempPrices[i + 1]; + } + } + } + // } +} + + + +static void LzmaEnc_Construct(CLzmaEnc *p) +{ + RangeEnc_Construct(&p->rc); + MatchFinder_Construct(&MFB); + + #ifndef _7ZIP_ST + p->matchFinderMt.MatchFinder = &MFB; + MatchFinderMt_Construct(&p->matchFinderMt); + #endif + + { + CLzmaEncProps props; + LzmaEncProps_Init(&props); + LzmaEnc_SetProps(p, &props); + } + + #ifndef LZMA_LOG_BSR + LzmaEnc_FastPosInit(p->g_FastPos); + #endif + + LzmaEnc_InitPriceTables(p->ProbPrices); + p->litProbs = NULL; + p->saveState.litProbs = NULL; +} + +CLzmaEncHandle LzmaEnc_Create(ISzAllocPtr alloc) +{ + void *p; + p = ISzAlloc_Alloc(alloc, sizeof(CLzmaEnc)); + if (p) + LzmaEnc_Construct((CLzmaEnc *)p); + return p; +} + +static void LzmaEnc_FreeLits(CLzmaEnc *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->litProbs); + ISzAlloc_Free(alloc, p->saveState.litProbs); + p->litProbs = NULL; + p->saveState.litProbs = NULL; +} + +static void LzmaEnc_Destruct(CLzmaEnc *p, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + #ifndef _7ZIP_ST + MatchFinderMt_Destruct(&p->matchFinderMt, allocBig); + #endif + + MatchFinder_Free(&MFB, allocBig); + LzmaEnc_FreeLits(p, alloc); + RangeEnc_Free(&p->rc, alloc); +} + +void LzmaEnc_Destroy(CLzmaEncHandle p, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + LzmaEnc_Destruct((CLzmaEnc *)p, alloc, allocBig); + ISzAlloc_Free(alloc, p); +} + + +MY_NO_INLINE +static SRes LzmaEnc_CodeOneBlock(CLzmaEnc *p, UInt32 maxPackSize, UInt32 maxUnpackSize) +{ + UInt32 nowPos32, startPos32; + if (p->needInit) + { + #ifndef _7ZIP_ST + if (p->mtMode) + { + RINOK(MatchFinderMt_InitMt(&p->matchFinderMt)); + } + #endif + p->matchFinder.Init(p->matchFinderObj); + p->needInit = 0; + } + + if (p->finished) + return p->result; + RINOK(CheckErrors(p)); + + nowPos32 = (UInt32)p->nowPos64; + startPos32 = nowPos32; + + if (p->nowPos64 == 0) + { + unsigned numPairs; + Byte curByte; + if (p->matchFinder.GetNumAvailableBytes(p->matchFinderObj) == 0) + return Flush(p, nowPos32); + ReadMatchDistances(p, &numPairs); + RangeEnc_EncodeBit_0(&p->rc, &p->isMatch[kState_Start][0]); + // p->state = kLiteralNextStates[p->state]; + curByte = *(p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - p->additionalOffset); + LitEnc_Encode(&p->rc, p->litProbs, curByte); + p->additionalOffset--; + nowPos32++; + } + + if (p->matchFinder.GetNumAvailableBytes(p->matchFinderObj) != 0) + + for (;;) + { + UInt32 dist; + unsigned len, posState; + UInt32 range, ttt, newBound; + CLzmaProb *probs; + + if (p->fastMode) + len = GetOptimumFast(p); + else + { + unsigned oci = p->optCur; + if (p->optEnd == oci) + len = GetOptimum(p, nowPos32); + else + { + const COptimal *opt = &p->opt[oci]; + len = opt->len; + p->backRes = opt->dist; + p->optCur = oci + 1; + } + } + + posState = (unsigned)nowPos32 & p->pbMask; + range = p->rc.range; + probs = &p->isMatch[p->state][posState]; + + RC_BIT_PRE(&p->rc, probs) + + dist = p->backRes; + + #ifdef SHOW_STAT2 + printf("\n pos = %6X, len = %3u pos = %6u", nowPos32, len, dist); + #endif + + if (dist == MARK_LIT) + { + Byte curByte; + const Byte *data; + unsigned state; + + RC_BIT_0(&p->rc, probs); + p->rc.range = range; + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - p->additionalOffset; + probs = LIT_PROBS(nowPos32, *(data - 1)); + curByte = *data; + state = p->state; + p->state = kLiteralNextStates[state]; + if (IsLitState(state)) + LitEnc_Encode(&p->rc, probs, curByte); + else + LitEnc_EncodeMatched(&p->rc, probs, curByte, *(data - p->reps[0])); + } + else + { + RC_BIT_1(&p->rc, probs); + probs = &p->isRep[p->state]; + RC_BIT_PRE(&p->rc, probs) + + if (dist < LZMA_NUM_REPS) + { + RC_BIT_1(&p->rc, probs); + probs = &p->isRepG0[p->state]; + RC_BIT_PRE(&p->rc, probs) + if (dist == 0) + { + RC_BIT_0(&p->rc, probs); + probs = &p->isRep0Long[p->state][posState]; + RC_BIT_PRE(&p->rc, probs) + if (len != 1) + { + RC_BIT_1_BASE(&p->rc, probs); + } + else + { + RC_BIT_0_BASE(&p->rc, probs); + p->state = kShortRepNextStates[p->state]; + } + } + else + { + RC_BIT_1(&p->rc, probs); + probs = &p->isRepG1[p->state]; + RC_BIT_PRE(&p->rc, probs) + if (dist == 1) + { + RC_BIT_0_BASE(&p->rc, probs); + dist = p->reps[1]; + } + else + { + RC_BIT_1(&p->rc, probs); + probs = &p->isRepG2[p->state]; + RC_BIT_PRE(&p->rc, probs) + if (dist == 2) + { + RC_BIT_0_BASE(&p->rc, probs); + dist = p->reps[2]; + } + else + { + RC_BIT_1_BASE(&p->rc, probs); + dist = p->reps[3]; + p->reps[3] = p->reps[2]; + } + p->reps[2] = p->reps[1]; + } + p->reps[1] = p->reps[0]; + p->reps[0] = dist; + } + + RC_NORM(&p->rc) + + p->rc.range = range; + + if (len != 1) + { + LenEnc_Encode(&p->repLenProbs, &p->rc, len - LZMA_MATCH_LEN_MIN, posState); + --p->repLenEncCounter; + p->state = kRepNextStates[p->state]; + } + } + else + { + unsigned posSlot; + RC_BIT_0(&p->rc, probs); + p->rc.range = range; + p->state = kMatchNextStates[p->state]; + + LenEnc_Encode(&p->lenProbs, &p->rc, len - LZMA_MATCH_LEN_MIN, posState); + // --p->lenEnc.counter; + + dist -= LZMA_NUM_REPS; + p->reps[3] = p->reps[2]; + p->reps[2] = p->reps[1]; + p->reps[1] = p->reps[0]; + p->reps[0] = dist + 1; + + p->matchPriceCount++; + GetPosSlot(dist, posSlot); + // RcTree_Encode_PosSlot(&p->rc, p->posSlotEncoder[GetLenToPosState(len)], posSlot); + { + UInt32 sym = (UInt32)posSlot + (1 << kNumPosSlotBits); + range = p->rc.range; + probs = p->posSlotEncoder[GetLenToPosState(len)]; + do + { + CLzmaProb *prob = probs + (sym >> kNumPosSlotBits); + UInt32 bit = (sym >> (kNumPosSlotBits - 1)) & 1; + sym <<= 1; + RC_BIT(&p->rc, prob, bit); + } + while (sym < (1 << kNumPosSlotBits * 2)); + p->rc.range = range; + } + + if (dist >= kStartPosModelIndex) + { + unsigned footerBits = ((posSlot >> 1) - 1); + + if (dist < kNumFullDistances) + { + unsigned base = ((2 | (posSlot & 1)) << footerBits); + RcTree_ReverseEncode(&p->rc, p->posEncoders + base, footerBits, (unsigned)(dist /* - base */)); + } + else + { + UInt32 pos2 = (dist | 0xF) << (32 - footerBits); + range = p->rc.range; + // RangeEnc_EncodeDirectBits(&p->rc, posReduced >> kNumAlignBits, footerBits - kNumAlignBits); + /* + do + { + range >>= 1; + p->rc.low += range & (0 - ((dist >> --footerBits) & 1)); + RC_NORM(&p->rc) + } + while (footerBits > kNumAlignBits); + */ + do + { + range >>= 1; + p->rc.low += range & (0 - (pos2 >> 31)); + pos2 += pos2; + RC_NORM(&p->rc) + } + while (pos2 != 0xF0000000); + + + // RcTree_ReverseEncode(&p->rc, p->posAlignEncoder, kNumAlignBits, posReduced & kAlignMask); + + { + unsigned m = 1; + unsigned bit; + bit = dist & 1; dist >>= 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); m = (m << 1) + bit; + bit = dist & 1; dist >>= 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); m = (m << 1) + bit; + bit = dist & 1; dist >>= 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); m = (m << 1) + bit; + bit = dist & 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); + p->rc.range = range; + // p->alignPriceCount++; + } + } + } + } + } + + nowPos32 += (UInt32)len; + p->additionalOffset -= len; + + if (p->additionalOffset == 0) + { + UInt32 processed; + + if (!p->fastMode) + { + /* + if (p->alignPriceCount >= 16) // kAlignTableSize + FillAlignPrices(p); + if (p->matchPriceCount >= 128) + FillDistancesPrices(p); + if (p->lenEnc.counter <= 0) + LenPriceEnc_UpdateTables(&p->lenEnc, 1 << p->pb, &p->lenProbs, p->ProbPrices); + */ + if (p->matchPriceCount >= 64) + { + FillAlignPrices(p); + // { int y; for (y = 0; y < 100; y++) { + FillDistancesPrices(p); + // }} + LenPriceEnc_UpdateTables(&p->lenEnc, (unsigned)1 << p->pb, &p->lenProbs, p->ProbPrices); + } + if (p->repLenEncCounter <= 0) + { + p->repLenEncCounter = REP_LEN_COUNT; + LenPriceEnc_UpdateTables(&p->repLenEnc, (unsigned)1 << p->pb, &p->repLenProbs, p->ProbPrices); + } + } + + if (p->matchFinder.GetNumAvailableBytes(p->matchFinderObj) == 0) + break; + processed = nowPos32 - startPos32; + + if (maxPackSize) + { + if (processed + kNumOpts + 300 >= maxUnpackSize + || RangeEnc_GetProcessed_sizet(&p->rc) + kPackReserve >= maxPackSize) + break; + } + else if (processed >= (1 << 17)) + { + p->nowPos64 += nowPos32 - startPos32; + return CheckErrors(p); + } + } + } + + p->nowPos64 += nowPos32 - startPos32; + return Flush(p, nowPos32); +} + + + +#define kBigHashDicLimit ((UInt32)1 << 24) + +static SRes LzmaEnc_Alloc(CLzmaEnc *p, UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + UInt32 beforeSize = kNumOpts; + UInt32 dictSize; + + if (!RangeEnc_Alloc(&p->rc, alloc)) + return SZ_ERROR_MEM; + + #ifndef _7ZIP_ST + p->mtMode = (p->multiThread && !p->fastMode && (MFB.btMode != 0)); + #endif + + { + unsigned lclp = p->lc + p->lp; + if (!p->litProbs || !p->saveState.litProbs || p->lclp != lclp) + { + LzmaEnc_FreeLits(p, alloc); + p->litProbs = (CLzmaProb *)ISzAlloc_Alloc(alloc, ((UInt32)0x300 << lclp) * sizeof(CLzmaProb)); + p->saveState.litProbs = (CLzmaProb *)ISzAlloc_Alloc(alloc, ((UInt32)0x300 << lclp) * sizeof(CLzmaProb)); + if (!p->litProbs || !p->saveState.litProbs) + { + LzmaEnc_FreeLits(p, alloc); + return SZ_ERROR_MEM; + } + p->lclp = lclp; + } + } + + MFB.bigHash = (Byte)(p->dictSize > kBigHashDicLimit ? 1 : 0); + + + dictSize = p->dictSize; + if (dictSize == ((UInt32)2 << 30) || + dictSize == ((UInt32)3 << 30)) + { + /* 21.03 : here we reduce the dictionary for 2 reasons: + 1) we don't want 32-bit back_distance matches in decoder for 2 GB dictionary. + 2) we want to elimate useless last MatchFinder_Normalize3() for corner cases, + where data size is aligned for 1 GB: 5/6/8 GB. + That reducing must be >= 1 for such corner cases. */ + dictSize -= 1; + } + + if (beforeSize + dictSize < keepWindowSize) + beforeSize = keepWindowSize - dictSize; + + /* in worst case we can look ahead for + max(LZMA_MATCH_LEN_MAX, numFastBytes + 1 + numFastBytes) bytes. + we send larger value for (keepAfter) to MantchFinder_Create(): + (numFastBytes + LZMA_MATCH_LEN_MAX + 1) + */ + + #ifndef _7ZIP_ST + if (p->mtMode) + { + RINOK(MatchFinderMt_Create(&p->matchFinderMt, dictSize, beforeSize, + p->numFastBytes, LZMA_MATCH_LEN_MAX + 1 /* 18.04 */ + , allocBig)); + p->matchFinderObj = &p->matchFinderMt; + MFB.bigHash = (Byte)( + (p->dictSize > kBigHashDicLimit && MFB.hashMask >= 0xFFFFFF) ? 1 : 0); + MatchFinderMt_CreateVTable(&p->matchFinderMt, &p->matchFinder); + } + else + #endif + { + if (!MatchFinder_Create(&MFB, dictSize, beforeSize, + p->numFastBytes, LZMA_MATCH_LEN_MAX + 1 /* 21.03 */ + , allocBig)) + return SZ_ERROR_MEM; + p->matchFinderObj = &MFB; + MatchFinder_CreateVTable(&MFB, &p->matchFinder); + } + + return SZ_OK; +} + +static void LzmaEnc_Init(CLzmaEnc *p) +{ + unsigned i; + p->state = 0; + p->reps[0] = + p->reps[1] = + p->reps[2] = + p->reps[3] = 1; + + RangeEnc_Init(&p->rc); + + for (i = 0; i < (1 << kNumAlignBits); i++) + p->posAlignEncoder[i] = kProbInitValue; + + for (i = 0; i < kNumStates; i++) + { + unsigned j; + for (j = 0; j < LZMA_NUM_PB_STATES_MAX; j++) + { + p->isMatch[i][j] = kProbInitValue; + p->isRep0Long[i][j] = kProbInitValue; + } + p->isRep[i] = kProbInitValue; + p->isRepG0[i] = kProbInitValue; + p->isRepG1[i] = kProbInitValue; + p->isRepG2[i] = kProbInitValue; + } + + { + for (i = 0; i < kNumLenToPosStates; i++) + { + CLzmaProb *probs = p->posSlotEncoder[i]; + unsigned j; + for (j = 0; j < (1 << kNumPosSlotBits); j++) + probs[j] = kProbInitValue; + } + } + { + for (i = 0; i < kNumFullDistances; i++) + p->posEncoders[i] = kProbInitValue; + } + + { + UInt32 num = (UInt32)0x300 << (p->lp + p->lc); + UInt32 k; + CLzmaProb *probs = p->litProbs; + for (k = 0; k < num; k++) + probs[k] = kProbInitValue; + } + + + LenEnc_Init(&p->lenProbs); + LenEnc_Init(&p->repLenProbs); + + p->optEnd = 0; + p->optCur = 0; + + { + for (i = 0; i < kNumOpts; i++) + p->opt[i].price = kInfinityPrice; + } + + p->additionalOffset = 0; + + p->pbMask = ((unsigned)1 << p->pb) - 1; + p->lpMask = ((UInt32)0x100 << p->lp) - ((unsigned)0x100 >> p->lc); + + // p->mf_Failure = False; +} + + +static void LzmaEnc_InitPrices(CLzmaEnc *p) +{ + if (!p->fastMode) + { + FillDistancesPrices(p); + FillAlignPrices(p); + } + + p->lenEnc.tableSize = + p->repLenEnc.tableSize = + p->numFastBytes + 1 - LZMA_MATCH_LEN_MIN; + + p->repLenEncCounter = REP_LEN_COUNT; + + LenPriceEnc_UpdateTables(&p->lenEnc, (unsigned)1 << p->pb, &p->lenProbs, p->ProbPrices); + LenPriceEnc_UpdateTables(&p->repLenEnc, (unsigned)1 << p->pb, &p->repLenProbs, p->ProbPrices); +} + +static SRes LzmaEnc_AllocAndInit(CLzmaEnc *p, UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + unsigned i; + for (i = kEndPosModelIndex / 2; i < kDicLogSizeMax; i++) + if (p->dictSize <= ((UInt32)1 << i)) + break; + p->distTableSize = i * 2; + + p->finished = False; + p->result = SZ_OK; + RINOK(LzmaEnc_Alloc(p, keepWindowSize, alloc, allocBig)); + LzmaEnc_Init(p); + LzmaEnc_InitPrices(p); + p->nowPos64 = 0; + return SZ_OK; +} + +static SRes LzmaEnc_Prepare(CLzmaEncHandle pp, ISeqOutStream *outStream, ISeqInStream *inStream, + ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + MFB.stream = inStream; + p->needInit = 1; + p->rc.outStream = outStream; + return LzmaEnc_AllocAndInit(p, 0, alloc, allocBig); +} + +SRes LzmaEnc_PrepareForLzma2(CLzmaEncHandle pp, + ISeqInStream *inStream, UInt32 keepWindowSize, + ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + MFB.stream = inStream; + p->needInit = 1; + return LzmaEnc_AllocAndInit(p, keepWindowSize, alloc, allocBig); +} + +static void LzmaEnc_SetInputBuf(CLzmaEnc *p, const Byte *src, SizeT srcLen) +{ + MFB.directInput = 1; + MFB.bufferBase = (Byte *)src; + MFB.directInputRem = srcLen; +} + +SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen, + UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + LzmaEnc_SetInputBuf(p, src, srcLen); + p->needInit = 1; + + LzmaEnc_SetDataSize(pp, srcLen); + return LzmaEnc_AllocAndInit(p, keepWindowSize, alloc, allocBig); +} + +void LzmaEnc_Finish(CLzmaEncHandle pp) +{ + #ifndef _7ZIP_ST + CLzmaEnc *p = (CLzmaEnc *)pp; + if (p->mtMode) + MatchFinderMt_ReleaseStream(&p->matchFinderMt); + #else + UNUSED_VAR(pp); + #endif +} + + +typedef struct +{ + ISeqOutStream vt; + Byte *data; + SizeT rem; + BoolInt overflow; +} CLzmaEnc_SeqOutStreamBuf; + +static size_t SeqOutStreamBuf_Write(const ISeqOutStream *pp, const void *data, size_t size) +{ + CLzmaEnc_SeqOutStreamBuf *p = CONTAINER_FROM_VTBL(pp, CLzmaEnc_SeqOutStreamBuf, vt); + if (p->rem < size) + { + size = p->rem; + p->overflow = True; + } + if (size != 0) + { + memcpy(p->data, data, size); + p->rem -= size; + p->data += size; + } + return size; +} + + +/* +UInt32 LzmaEnc_GetNumAvailableBytes(CLzmaEncHandle pp) +{ + const CLzmaEnc *p = (CLzmaEnc *)pp; + return p->matchFinder.GetNumAvailableBytes(p->matchFinderObj); +} +*/ + +const Byte *LzmaEnc_GetCurBuf(CLzmaEncHandle pp) +{ + const CLzmaEnc *p = (CLzmaEnc *)pp; + return p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - p->additionalOffset; +} + + +// (desiredPackSize == 0) is not allowed +SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit, + Byte *dest, size_t *destLen, UInt32 desiredPackSize, UInt32 *unpackSize) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + UInt64 nowPos64; + SRes res; + CLzmaEnc_SeqOutStreamBuf outStream; + + outStream.vt.Write = SeqOutStreamBuf_Write; + outStream.data = dest; + outStream.rem = *destLen; + outStream.overflow = False; + + p->writeEndMark = False; + p->finished = False; + p->result = SZ_OK; + + if (reInit) + LzmaEnc_Init(p); + LzmaEnc_InitPrices(p); + RangeEnc_Init(&p->rc); + p->rc.outStream = &outStream.vt; + nowPos64 = p->nowPos64; + + res = LzmaEnc_CodeOneBlock(p, desiredPackSize, *unpackSize); + + *unpackSize = (UInt32)(p->nowPos64 - nowPos64); + *destLen -= outStream.rem; + if (outStream.overflow) + return SZ_ERROR_OUTPUT_EOF; + + return res; +} + + +MY_NO_INLINE +static SRes LzmaEnc_Encode2(CLzmaEnc *p, ICompressProgress *progress) +{ + SRes res = SZ_OK; + + #ifndef _7ZIP_ST + Byte allocaDummy[0x300]; + allocaDummy[0] = 0; + allocaDummy[1] = allocaDummy[0]; + #endif + + for (;;) + { + res = LzmaEnc_CodeOneBlock(p, 0, 0); + if (res != SZ_OK || p->finished) + break; + if (progress) + { + res = ICompressProgress_Progress(progress, p->nowPos64, RangeEnc_GetProcessed(&p->rc)); + if (res != SZ_OK) + { + res = SZ_ERROR_PROGRESS; + break; + } + } + } + + LzmaEnc_Finish(p); + + /* + if (res == SZ_OK && !Inline_MatchFinder_IsFinishedOK(&MFB)) + res = SZ_ERROR_FAIL; + } + */ + + return res; +} + + +SRes LzmaEnc_Encode(CLzmaEncHandle pp, ISeqOutStream *outStream, ISeqInStream *inStream, ICompressProgress *progress, + ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + RINOK(LzmaEnc_Prepare(pp, outStream, inStream, alloc, allocBig)); + return LzmaEnc_Encode2((CLzmaEnc *)pp, progress); +} + + +SRes LzmaEnc_WriteProperties(CLzmaEncHandle pp, Byte *props, SizeT *size) +{ + if (*size < LZMA_PROPS_SIZE) + return SZ_ERROR_PARAM; + *size = LZMA_PROPS_SIZE; + { + const CLzmaEnc *p = (const CLzmaEnc *)pp; + const UInt32 dictSize = p->dictSize; + UInt32 v; + props[0] = (Byte)((p->pb * 5 + p->lp) * 9 + p->lc); + + // we write aligned dictionary value to properties for lzma decoder + if (dictSize >= ((UInt32)1 << 21)) + { + const UInt32 kDictMask = ((UInt32)1 << 20) - 1; + v = (dictSize + kDictMask) & ~kDictMask; + if (v < dictSize) + v = dictSize; + } + else + { + unsigned i = 11 * 2; + do + { + v = (UInt32)(2 + (i & 1)) << (i >> 1); + i++; + } + while (v < dictSize); + } + + SetUi32(props + 1, v); + return SZ_OK; + } +} + + +unsigned LzmaEnc_IsWriteEndMark(CLzmaEncHandle pp) +{ + return (unsigned)((CLzmaEnc *)pp)->writeEndMark; +} + + +SRes LzmaEnc_MemEncode(CLzmaEncHandle pp, Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + int writeEndMark, ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + SRes res; + CLzmaEnc *p = (CLzmaEnc *)pp; + + CLzmaEnc_SeqOutStreamBuf outStream; + + outStream.vt.Write = SeqOutStreamBuf_Write; + outStream.data = dest; + outStream.rem = *destLen; + outStream.overflow = False; + + p->writeEndMark = writeEndMark; + p->rc.outStream = &outStream.vt; + + res = LzmaEnc_MemPrepare(pp, src, srcLen, 0, alloc, allocBig); + + if (res == SZ_OK) + { + res = LzmaEnc_Encode2(p, progress); + if (res == SZ_OK && p->nowPos64 != srcLen) + res = SZ_ERROR_FAIL; + } + + *destLen -= outStream.rem; + if (outStream.overflow) + return SZ_ERROR_OUTPUT_EOF; + return res; +} + + +SRes LzmaEncode(Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + const CLzmaEncProps *props, Byte *propsEncoded, SizeT *propsSize, int writeEndMark, + ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)LzmaEnc_Create(alloc); + SRes res; + if (!p) + return SZ_ERROR_MEM; + + res = LzmaEnc_SetProps(p, props); + if (res == SZ_OK) + { + res = LzmaEnc_WriteProperties(p, propsEncoded, propsSize); + if (res == SZ_OK) + res = LzmaEnc_MemEncode(p, dest, destLen, src, srcLen, + writeEndMark, progress, alloc, allocBig); + } + + LzmaEnc_Destroy(p, alloc, allocBig); + return res; +} + + +/* +#ifndef _7ZIP_ST +void LzmaEnc_GetLzThreads(CLzmaEncHandle pp, HANDLE lz_threads[2]) +{ + const CLzmaEnc *p = (CLzmaEnc *)pp; + lz_threads[0] = p->matchFinderMt.hashSync.thread; + lz_threads[1] = p->matchFinderMt.btSync.thread; +} +#endif +*/ diff --git a/deps/LZMA-SDK/C/LzmaEnc.h b/deps/LZMA-SDK/C/LzmaEnc.h index 26757ba6b..bc2ed5042 100644 --- a/deps/LZMA-SDK/C/LzmaEnc.h +++ b/deps/LZMA-SDK/C/LzmaEnc.h @@ -1,78 +1,78 @@ -/* LzmaEnc.h -- LZMA Encoder -2019-10-30 : Igor Pavlov : Public domain */ - -#ifndef __LZMA_ENC_H -#define __LZMA_ENC_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -#define LZMA_PROPS_SIZE 5 - -typedef struct _CLzmaEncProps -{ - int level; /* 0 <= level <= 9 */ - UInt32 dictSize; /* (1 << 12) <= dictSize <= (1 << 27) for 32-bit version - (1 << 12) <= dictSize <= (3 << 29) for 64-bit version - default = (1 << 24) */ - int lc; /* 0 <= lc <= 8, default = 3 */ - int lp; /* 0 <= lp <= 4, default = 0 */ - int pb; /* 0 <= pb <= 4, default = 2 */ - int algo; /* 0 - fast, 1 - normal, default = 1 */ - int fb; /* 5 <= fb <= 273, default = 32 */ - int btMode; /* 0 - hashChain Mode, 1 - binTree mode - normal, default = 1 */ - int numHashBytes; /* 2, 3 or 4, default = 4 */ - UInt32 mc; /* 1 <= mc <= (1 << 30), default = 32 */ - unsigned writeEndMark; /* 0 - do not write EOPM, 1 - write EOPM, default = 0 */ - int numThreads; /* 1 or 2, default = 2 */ - - UInt64 reduceSize; /* estimated size of data that will be compressed. default = (UInt64)(Int64)-1. - Encoder uses this value to reduce dictionary size */ - - UInt64 affinity; -} CLzmaEncProps; - -void LzmaEncProps_Init(CLzmaEncProps *p); -void LzmaEncProps_Normalize(CLzmaEncProps *p); -UInt32 LzmaEncProps_GetDictSize(const CLzmaEncProps *props2); - - -/* ---------- CLzmaEncHandle Interface ---------- */ - -/* LzmaEnc* functions can return the following exit codes: -SRes: - SZ_OK - OK - SZ_ERROR_MEM - Memory allocation error - SZ_ERROR_PARAM - Incorrect paramater in props - SZ_ERROR_WRITE - ISeqOutStream write callback error - SZ_ERROR_OUTPUT_EOF - output buffer overflow - version with (Byte *) output - SZ_ERROR_PROGRESS - some break from progress callback - SZ_ERROR_THREAD - error in multithreading functions (only for Mt version) -*/ - -typedef void * CLzmaEncHandle; - -CLzmaEncHandle LzmaEnc_Create(ISzAllocPtr alloc); -void LzmaEnc_Destroy(CLzmaEncHandle p, ISzAllocPtr alloc, ISzAllocPtr allocBig); - -SRes LzmaEnc_SetProps(CLzmaEncHandle p, const CLzmaEncProps *props); -void LzmaEnc_SetDataSize(CLzmaEncHandle p, UInt64 expectedDataSiize); -SRes LzmaEnc_WriteProperties(CLzmaEncHandle p, Byte *properties, SizeT *size); -unsigned LzmaEnc_IsWriteEndMark(CLzmaEncHandle p); - -SRes LzmaEnc_Encode(CLzmaEncHandle p, ISeqOutStream *outStream, ISeqInStream *inStream, - ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig); -SRes LzmaEnc_MemEncode(CLzmaEncHandle p, Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, - int writeEndMark, ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig); - - -/* ---------- One Call Interface ---------- */ - -SRes LzmaEncode(Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, - const CLzmaEncProps *props, Byte *propsEncoded, SizeT *propsSize, int writeEndMark, - ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig); - -EXTERN_C_END - -#endif +/* LzmaEnc.h -- LZMA Encoder +2019-10-30 : Igor Pavlov : Public domain */ + +#ifndef __LZMA_ENC_H +#define __LZMA_ENC_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define LZMA_PROPS_SIZE 5 + +typedef struct _CLzmaEncProps +{ + int level; /* 0 <= level <= 9 */ + UInt32 dictSize; /* (1 << 12) <= dictSize <= (1 << 27) for 32-bit version + (1 << 12) <= dictSize <= (3 << 29) for 64-bit version + default = (1 << 24) */ + int lc; /* 0 <= lc <= 8, default = 3 */ + int lp; /* 0 <= lp <= 4, default = 0 */ + int pb; /* 0 <= pb <= 4, default = 2 */ + int algo; /* 0 - fast, 1 - normal, default = 1 */ + int fb; /* 5 <= fb <= 273, default = 32 */ + int btMode; /* 0 - hashChain Mode, 1 - binTree mode - normal, default = 1 */ + int numHashBytes; /* 2, 3 or 4, default = 4 */ + UInt32 mc; /* 1 <= mc <= (1 << 30), default = 32 */ + unsigned writeEndMark; /* 0 - do not write EOPM, 1 - write EOPM, default = 0 */ + int numThreads; /* 1 or 2, default = 2 */ + + UInt64 reduceSize; /* estimated size of data that will be compressed. default = (UInt64)(Int64)-1. + Encoder uses this value to reduce dictionary size */ + + UInt64 affinity; +} CLzmaEncProps; + +void LzmaEncProps_Init(CLzmaEncProps *p); +void LzmaEncProps_Normalize(CLzmaEncProps *p); +UInt32 LzmaEncProps_GetDictSize(const CLzmaEncProps *props2); + + +/* ---------- CLzmaEncHandle Interface ---------- */ + +/* LzmaEnc* functions can return the following exit codes: +SRes: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater in props + SZ_ERROR_WRITE - ISeqOutStream write callback error + SZ_ERROR_OUTPUT_EOF - output buffer overflow - version with (Byte *) output + SZ_ERROR_PROGRESS - some break from progress callback + SZ_ERROR_THREAD - error in multithreading functions (only for Mt version) +*/ + +typedef void * CLzmaEncHandle; + +CLzmaEncHandle LzmaEnc_Create(ISzAllocPtr alloc); +void LzmaEnc_Destroy(CLzmaEncHandle p, ISzAllocPtr alloc, ISzAllocPtr allocBig); + +SRes LzmaEnc_SetProps(CLzmaEncHandle p, const CLzmaEncProps *props); +void LzmaEnc_SetDataSize(CLzmaEncHandle p, UInt64 expectedDataSiize); +SRes LzmaEnc_WriteProperties(CLzmaEncHandle p, Byte *properties, SizeT *size); +unsigned LzmaEnc_IsWriteEndMark(CLzmaEncHandle p); + +SRes LzmaEnc_Encode(CLzmaEncHandle p, ISeqOutStream *outStream, ISeqInStream *inStream, + ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig); +SRes LzmaEnc_MemEncode(CLzmaEncHandle p, Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + int writeEndMark, ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig); + + +/* ---------- One Call Interface ---------- */ + +SRes LzmaEncode(Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + const CLzmaEncProps *props, Byte *propsEncoded, SizeT *propsSize, int writeEndMark, + ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/LzmaLib.c b/deps/LZMA-SDK/C/LzmaLib.c index c10cf1a0f..706e9e58c 100644 --- a/deps/LZMA-SDK/C/LzmaLib.c +++ b/deps/LZMA-SDK/C/LzmaLib.c @@ -1,40 +1,40 @@ -/* LzmaLib.c -- LZMA library wrapper -2015-06-13 : Igor Pavlov : Public domain */ - -#include "Alloc.h" -#include "LzmaDec.h" -#include "LzmaEnc.h" -#include "LzmaLib.h" - -MY_STDAPI LzmaCompress(unsigned char *dest, size_t *destLen, const unsigned char *src, size_t srcLen, - unsigned char *outProps, size_t *outPropsSize, - int level, /* 0 <= level <= 9, default = 5 */ - unsigned dictSize, /* use (1 << N) or (3 << N). 4 KB < dictSize <= 128 MB */ - int lc, /* 0 <= lc <= 8, default = 3 */ - int lp, /* 0 <= lp <= 4, default = 0 */ - int pb, /* 0 <= pb <= 4, default = 2 */ - int fb, /* 5 <= fb <= 273, default = 32 */ - int numThreads /* 1 or 2, default = 2 */ -) -{ - CLzmaEncProps props; - LzmaEncProps_Init(&props); - props.level = level; - props.dictSize = dictSize; - props.lc = lc; - props.lp = lp; - props.pb = pb; - props.fb = fb; - props.numThreads = numThreads; - - return LzmaEncode(dest, destLen, src, srcLen, &props, outProps, outPropsSize, 0, - NULL, &g_Alloc, &g_Alloc); -} - - -MY_STDAPI LzmaUncompress(unsigned char *dest, size_t *destLen, const unsigned char *src, size_t *srcLen, - const unsigned char *props, size_t propsSize) -{ - ELzmaStatus status; - return LzmaDecode(dest, destLen, src, srcLen, props, (unsigned)propsSize, LZMA_FINISH_ANY, &status, &g_Alloc); -} +/* LzmaLib.c -- LZMA library wrapper +2015-06-13 : Igor Pavlov : Public domain */ + +#include "Alloc.h" +#include "LzmaDec.h" +#include "LzmaEnc.h" +#include "LzmaLib.h" + +MY_STDAPI LzmaCompress(unsigned char *dest, size_t *destLen, const unsigned char *src, size_t srcLen, + unsigned char *outProps, size_t *outPropsSize, + int level, /* 0 <= level <= 9, default = 5 */ + unsigned dictSize, /* use (1 << N) or (3 << N). 4 KB < dictSize <= 128 MB */ + int lc, /* 0 <= lc <= 8, default = 3 */ + int lp, /* 0 <= lp <= 4, default = 0 */ + int pb, /* 0 <= pb <= 4, default = 2 */ + int fb, /* 5 <= fb <= 273, default = 32 */ + int numThreads /* 1 or 2, default = 2 */ +) +{ + CLzmaEncProps props; + LzmaEncProps_Init(&props); + props.level = level; + props.dictSize = dictSize; + props.lc = lc; + props.lp = lp; + props.pb = pb; + props.fb = fb; + props.numThreads = numThreads; + + return LzmaEncode(dest, destLen, src, srcLen, &props, outProps, outPropsSize, 0, + NULL, &g_Alloc, &g_Alloc); +} + + +MY_STDAPI LzmaUncompress(unsigned char *dest, size_t *destLen, const unsigned char *src, size_t *srcLen, + const unsigned char *props, size_t propsSize) +{ + ELzmaStatus status; + return LzmaDecode(dest, destLen, src, srcLen, props, (unsigned)propsSize, LZMA_FINISH_ANY, &status, &g_Alloc); +} diff --git a/deps/LZMA-SDK/C/LzmaLib.h b/deps/LZMA-SDK/C/LzmaLib.h index 4103e224a..c343a8596 100644 --- a/deps/LZMA-SDK/C/LzmaLib.h +++ b/deps/LZMA-SDK/C/LzmaLib.h @@ -1,138 +1,138 @@ -/* LzmaLib.h -- LZMA library interface -2021-04-03 : Igor Pavlov : Public domain */ - -#ifndef __LZMA_LIB_H -#define __LZMA_LIB_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -#define MY_STDAPI int MY_STD_CALL - -#define LZMA_PROPS_SIZE 5 - -/* -RAM requirements for LZMA: - for compression: (dictSize * 11.5 + 6 MB) + state_size - for decompression: dictSize + state_size - state_size = (4 + (1.5 << (lc + lp))) KB - by default (lc=3, lp=0), state_size = 16 KB. - -LZMA properties (5 bytes) format - Offset Size Description - 0 1 lc, lp and pb in encoded form. - 1 4 dictSize (little endian). -*/ - -/* -LzmaCompress ------------- - -outPropsSize - - In: the pointer to the size of outProps buffer; *outPropsSize = LZMA_PROPS_SIZE = 5. - Out: the pointer to the size of written properties in outProps buffer; *outPropsSize = LZMA_PROPS_SIZE = 5. - - LZMA Encoder will use defult values for any parameter, if it is - -1 for any from: level, loc, lp, pb, fb, numThreads - 0 for dictSize - -level - compression level: 0 <= level <= 9; - - level dictSize algo fb - 0: 64 KB 0 32 - 1: 256 KB 0 32 - 2: 1 MB 0 32 - 3: 4 MB 0 32 - 4: 16 MB 0 32 - 5: 16 MB 1 32 - 6: 32 MB 1 32 - 7: 32 MB 1 64 - 8: 64 MB 1 64 - 9: 64 MB 1 64 - - The default value for "level" is 5. - - algo = 0 means fast method - algo = 1 means normal method - -dictSize - The dictionary size in bytes. The maximum value is - 128 MB = (1 << 27) bytes for 32-bit version - 1 GB = (1 << 30) bytes for 64-bit version - The default value is 16 MB = (1 << 24) bytes. - It's recommended to use the dictionary that is larger than 4 KB and - that can be calculated as (1 << N) or (3 << N) sizes. - -lc - The number of literal context bits (high bits of previous literal). - It can be in the range from 0 to 8. The default value is 3. - Sometimes lc=4 gives the gain for big files. - -lp - The number of literal pos bits (low bits of current position for literals). - It can be in the range from 0 to 4. The default value is 0. - The lp switch is intended for periodical data when the period is equal to 2^lp. - For example, for 32-bit (4 bytes) periodical data you can use lp=2. Often it's - better to set lc=0, if you change lp switch. - -pb - The number of pos bits (low bits of current position). - It can be in the range from 0 to 4. The default value is 2. - The pb switch is intended for periodical data when the period is equal 2^pb. - -fb - Word size (the number of fast bytes). - It can be in the range from 5 to 273. The default value is 32. - Usually, a big number gives a little bit better compression ratio and - slower compression process. - -numThreads - The number of thereads. 1 or 2. The default value is 2. - Fast mode (algo = 0) can use only 1 thread. - -In: - dest - output data buffer - destLen - output data buffer size - src - input data - srcLen - input data size -Out: - destLen - processed output size -Returns: - SZ_OK - OK - SZ_ERROR_MEM - Memory allocation error - SZ_ERROR_PARAM - Incorrect paramater - SZ_ERROR_OUTPUT_EOF - output buffer overflow - SZ_ERROR_THREAD - errors in multithreading functions (only for Mt version) -*/ - -MY_STDAPI LzmaCompress(unsigned char *dest, size_t *destLen, const unsigned char *src, size_t srcLen, - unsigned char *outProps, size_t *outPropsSize, /* *outPropsSize must be = 5 */ - int level, /* 0 <= level <= 9, default = 5 */ - unsigned dictSize, /* default = (1 << 24) */ - int lc, /* 0 <= lc <= 8, default = 3 */ - int lp, /* 0 <= lp <= 4, default = 0 */ - int pb, /* 0 <= pb <= 4, default = 2 */ - int fb, /* 5 <= fb <= 273, default = 32 */ - int numThreads /* 1 or 2, default = 2 */ - ); - -/* -LzmaUncompress --------------- -In: - dest - output data buffer - destLen - output data buffer size - src - input data - srcLen - input data size -Out: - destLen - processed output size - srcLen - processed input size -Returns: - SZ_OK - OK - SZ_ERROR_DATA - Data error - SZ_ERROR_MEM - Memory allocation arror - SZ_ERROR_UNSUPPORTED - Unsupported properties - SZ_ERROR_INPUT_EOF - it needs more bytes in input buffer (src) -*/ - -MY_STDAPI LzmaUncompress(unsigned char *dest, size_t *destLen, const unsigned char *src, SizeT *srcLen, - const unsigned char *props, size_t propsSize); - -EXTERN_C_END - -#endif +/* LzmaLib.h -- LZMA library interface +2021-04-03 : Igor Pavlov : Public domain */ + +#ifndef __LZMA_LIB_H +#define __LZMA_LIB_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define MY_STDAPI int MY_STD_CALL + +#define LZMA_PROPS_SIZE 5 + +/* +RAM requirements for LZMA: + for compression: (dictSize * 11.5 + 6 MB) + state_size + for decompression: dictSize + state_size + state_size = (4 + (1.5 << (lc + lp))) KB + by default (lc=3, lp=0), state_size = 16 KB. + +LZMA properties (5 bytes) format + Offset Size Description + 0 1 lc, lp and pb in encoded form. + 1 4 dictSize (little endian). +*/ + +/* +LzmaCompress +------------ + +outPropsSize - + In: the pointer to the size of outProps buffer; *outPropsSize = LZMA_PROPS_SIZE = 5. + Out: the pointer to the size of written properties in outProps buffer; *outPropsSize = LZMA_PROPS_SIZE = 5. + + LZMA Encoder will use defult values for any parameter, if it is + -1 for any from: level, loc, lp, pb, fb, numThreads + 0 for dictSize + +level - compression level: 0 <= level <= 9; + + level dictSize algo fb + 0: 64 KB 0 32 + 1: 256 KB 0 32 + 2: 1 MB 0 32 + 3: 4 MB 0 32 + 4: 16 MB 0 32 + 5: 16 MB 1 32 + 6: 32 MB 1 32 + 7: 32 MB 1 64 + 8: 64 MB 1 64 + 9: 64 MB 1 64 + + The default value for "level" is 5. + + algo = 0 means fast method + algo = 1 means normal method + +dictSize - The dictionary size in bytes. The maximum value is + 128 MB = (1 << 27) bytes for 32-bit version + 1 GB = (1 << 30) bytes for 64-bit version + The default value is 16 MB = (1 << 24) bytes. + It's recommended to use the dictionary that is larger than 4 KB and + that can be calculated as (1 << N) or (3 << N) sizes. + +lc - The number of literal context bits (high bits of previous literal). + It can be in the range from 0 to 8. The default value is 3. + Sometimes lc=4 gives the gain for big files. + +lp - The number of literal pos bits (low bits of current position for literals). + It can be in the range from 0 to 4. The default value is 0. + The lp switch is intended for periodical data when the period is equal to 2^lp. + For example, for 32-bit (4 bytes) periodical data you can use lp=2. Often it's + better to set lc=0, if you change lp switch. + +pb - The number of pos bits (low bits of current position). + It can be in the range from 0 to 4. The default value is 2. + The pb switch is intended for periodical data when the period is equal 2^pb. + +fb - Word size (the number of fast bytes). + It can be in the range from 5 to 273. The default value is 32. + Usually, a big number gives a little bit better compression ratio and + slower compression process. + +numThreads - The number of thereads. 1 or 2. The default value is 2. + Fast mode (algo = 0) can use only 1 thread. + +In: + dest - output data buffer + destLen - output data buffer size + src - input data + srcLen - input data size +Out: + destLen - processed output size +Returns: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater + SZ_ERROR_OUTPUT_EOF - output buffer overflow + SZ_ERROR_THREAD - errors in multithreading functions (only for Mt version) +*/ + +MY_STDAPI LzmaCompress(unsigned char *dest, size_t *destLen, const unsigned char *src, size_t srcLen, + unsigned char *outProps, size_t *outPropsSize, /* *outPropsSize must be = 5 */ + int level, /* 0 <= level <= 9, default = 5 */ + unsigned dictSize, /* default = (1 << 24) */ + int lc, /* 0 <= lc <= 8, default = 3 */ + int lp, /* 0 <= lp <= 4, default = 0 */ + int pb, /* 0 <= pb <= 4, default = 2 */ + int fb, /* 5 <= fb <= 273, default = 32 */ + int numThreads /* 1 or 2, default = 2 */ + ); + +/* +LzmaUncompress +-------------- +In: + dest - output data buffer + destLen - output data buffer size + src - input data + srcLen - input data size +Out: + destLen - processed output size + srcLen - processed input size +Returns: + SZ_OK - OK + SZ_ERROR_DATA - Data error + SZ_ERROR_MEM - Memory allocation arror + SZ_ERROR_UNSUPPORTED - Unsupported properties + SZ_ERROR_INPUT_EOF - it needs more bytes in input buffer (src) +*/ + +MY_STDAPI LzmaUncompress(unsigned char *dest, size_t *destLen, const unsigned char *src, SizeT *srcLen, + const unsigned char *props, size_t propsSize); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/MtCoder.c b/deps/LZMA-SDK/C/MtCoder.c index e39d9cb19..99dc9090a 100644 --- a/deps/LZMA-SDK/C/MtCoder.c +++ b/deps/LZMA-SDK/C/MtCoder.c @@ -1,595 +1,595 @@ -/* MtCoder.c -- Multi-thread Coder -2021-12-21 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "MtCoder.h" - -#ifndef _7ZIP_ST - -static SRes MtProgressThunk_Progress(const ICompressProgress *pp, UInt64 inSize, UInt64 outSize) -{ - CMtProgressThunk *thunk = CONTAINER_FROM_VTBL(pp, CMtProgressThunk, vt); - UInt64 inSize2 = 0; - UInt64 outSize2 = 0; - if (inSize != (UInt64)(Int64)-1) - { - inSize2 = inSize - thunk->inSize; - thunk->inSize = inSize; - } - if (outSize != (UInt64)(Int64)-1) - { - outSize2 = outSize - thunk->outSize; - thunk->outSize = outSize; - } - return MtProgress_ProgressAdd(thunk->mtProgress, inSize2, outSize2); -} - - -void MtProgressThunk_CreateVTable(CMtProgressThunk *p) -{ - p->vt.Progress = MtProgressThunk_Progress; -} - - - -#define RINOK_THREAD(x) { if ((x) != 0) return SZ_ERROR_THREAD; } - - -static WRes ArEvent_OptCreate_And_Reset(CEvent *p) -{ - if (Event_IsCreated(p)) - return Event_Reset(p); - return AutoResetEvent_CreateNotSignaled(p); -} - - -static THREAD_FUNC_DECL ThreadFunc(void *pp); - - -static SRes MtCoderThread_CreateAndStart(CMtCoderThread *t) -{ - WRes wres = ArEvent_OptCreate_And_Reset(&t->startEvent); - if (wres == 0) - { - t->stop = False; - if (!Thread_WasCreated(&t->thread)) - wres = Thread_Create(&t->thread, ThreadFunc, t); - if (wres == 0) - wres = Event_Set(&t->startEvent); - } - if (wres == 0) - return SZ_OK; - return MY_SRes_HRESULT_FROM_WRes(wres); -} - - -static void MtCoderThread_Destruct(CMtCoderThread *t) -{ - if (Thread_WasCreated(&t->thread)) - { - t->stop = 1; - Event_Set(&t->startEvent); - Thread_Wait_Close(&t->thread); - } - - Event_Close(&t->startEvent); - - if (t->inBuf) - { - ISzAlloc_Free(t->mtCoder->allocBig, t->inBuf); - t->inBuf = NULL; - } -} - - - -static SRes FullRead(ISeqInStream *stream, Byte *data, size_t *processedSize) -{ - size_t size = *processedSize; - *processedSize = 0; - while (size != 0) - { - size_t cur = size; - SRes res = ISeqInStream_Read(stream, data, &cur); - *processedSize += cur; - data += cur; - size -= cur; - RINOK(res); - if (cur == 0) - return SZ_OK; - } - return SZ_OK; -} - - -/* - ThreadFunc2() returns: - SZ_OK - in all normal cases (even for stream error or memory allocation error) - SZ_ERROR_THREAD - in case of failure in system synch function -*/ - -static SRes ThreadFunc2(CMtCoderThread *t) -{ - CMtCoder *mtc = t->mtCoder; - - for (;;) - { - unsigned bi; - SRes res; - SRes res2; - BoolInt finished; - unsigned bufIndex; - size_t size; - const Byte *inData; - UInt64 readProcessed = 0; - - RINOK_THREAD(Event_Wait(&mtc->readEvent)) - - /* after Event_Wait(&mtc->readEvent) we must call Event_Set(&mtc->readEvent) in any case to unlock another threads */ - - if (mtc->stopReading) - { - return Event_Set(&mtc->readEvent) == 0 ? SZ_OK : SZ_ERROR_THREAD; - } - - res = MtProgress_GetError(&mtc->mtProgress); - - size = 0; - inData = NULL; - finished = True; - - if (res == SZ_OK) - { - size = mtc->blockSize; - if (mtc->inStream) - { - if (!t->inBuf) - { - t->inBuf = (Byte *)ISzAlloc_Alloc(mtc->allocBig, mtc->blockSize); - if (!t->inBuf) - res = SZ_ERROR_MEM; - } - if (res == SZ_OK) - { - res = FullRead(mtc->inStream, t->inBuf, &size); - readProcessed = mtc->readProcessed + size; - mtc->readProcessed = readProcessed; - } - if (res != SZ_OK) - { - mtc->readRes = res; - /* after reading error - we can stop encoding of previous blocks */ - MtProgress_SetError(&mtc->mtProgress, res); - } - else - finished = (size != mtc->blockSize); - } - else - { - size_t rem; - readProcessed = mtc->readProcessed; - rem = mtc->inDataSize - (size_t)readProcessed; - if (size > rem) - size = rem; - inData = mtc->inData + (size_t)readProcessed; - readProcessed += size; - mtc->readProcessed = readProcessed; - finished = (mtc->inDataSize == (size_t)readProcessed); - } - } - - /* we must get some block from blocksSemaphore before Event_Set(&mtc->readEvent) */ - - res2 = SZ_OK; - - if (Semaphore_Wait(&mtc->blocksSemaphore) != 0) - { - res2 = SZ_ERROR_THREAD; - if (res == SZ_OK) - { - res = res2; - // MtProgress_SetError(&mtc->mtProgress, res); - } - } - - bi = mtc->blockIndex; - - if (++mtc->blockIndex >= mtc->numBlocksMax) - mtc->blockIndex = 0; - - bufIndex = (unsigned)(int)-1; - - if (res == SZ_OK) - res = MtProgress_GetError(&mtc->mtProgress); - - if (res != SZ_OK) - finished = True; - - if (!finished) - { - if (mtc->numStartedThreads < mtc->numStartedThreadsLimit - && mtc->expectedDataSize != readProcessed) - { - res = MtCoderThread_CreateAndStart(&mtc->threads[mtc->numStartedThreads]); - if (res == SZ_OK) - mtc->numStartedThreads++; - else - { - MtProgress_SetError(&mtc->mtProgress, res); - finished = True; - } - } - } - - if (finished) - mtc->stopReading = True; - - RINOK_THREAD(Event_Set(&mtc->readEvent)) - - if (res2 != SZ_OK) - return res2; - - if (res == SZ_OK) - { - CriticalSection_Enter(&mtc->cs); - bufIndex = mtc->freeBlockHead; - mtc->freeBlockHead = mtc->freeBlockList[bufIndex]; - CriticalSection_Leave(&mtc->cs); - - res = mtc->mtCallback->Code(mtc->mtCallbackObject, t->index, bufIndex, - mtc->inStream ? t->inBuf : inData, size, finished); - - // MtProgress_Reinit(&mtc->mtProgress, t->index); - - if (res != SZ_OK) - MtProgress_SetError(&mtc->mtProgress, res); - } - - { - CMtCoderBlock *block = &mtc->blocks[bi]; - block->res = res; - block->bufIndex = bufIndex; - block->finished = finished; - } - - #ifdef MTCODER__USE_WRITE_THREAD - RINOK_THREAD(Event_Set(&mtc->writeEvents[bi])) - #else - { - unsigned wi; - { - CriticalSection_Enter(&mtc->cs); - wi = mtc->writeIndex; - if (wi == bi) - mtc->writeIndex = (unsigned)(int)-1; - else - mtc->ReadyBlocks[bi] = True; - CriticalSection_Leave(&mtc->cs); - } - - if (wi != bi) - { - if (res != SZ_OK || finished) - return 0; - continue; - } - - if (mtc->writeRes != SZ_OK) - res = mtc->writeRes; - - for (;;) - { - if (res == SZ_OK && bufIndex != (unsigned)(int)-1) - { - res = mtc->mtCallback->Write(mtc->mtCallbackObject, bufIndex); - if (res != SZ_OK) - { - mtc->writeRes = res; - MtProgress_SetError(&mtc->mtProgress, res); - } - } - - if (++wi >= mtc->numBlocksMax) - wi = 0; - { - BoolInt isReady; - - CriticalSection_Enter(&mtc->cs); - - if (bufIndex != (unsigned)(int)-1) - { - mtc->freeBlockList[bufIndex] = mtc->freeBlockHead; - mtc->freeBlockHead = bufIndex; - } - - isReady = mtc->ReadyBlocks[wi]; - - if (isReady) - mtc->ReadyBlocks[wi] = False; - else - mtc->writeIndex = wi; - - CriticalSection_Leave(&mtc->cs); - - RINOK_THREAD(Semaphore_Release1(&mtc->blocksSemaphore)) - - if (!isReady) - break; - } - - { - CMtCoderBlock *block = &mtc->blocks[wi]; - if (res == SZ_OK && block->res != SZ_OK) - res = block->res; - bufIndex = block->bufIndex; - finished = block->finished; - } - } - } - #endif - - if (finished || res != SZ_OK) - return 0; - } -} - - -static THREAD_FUNC_DECL ThreadFunc(void *pp) -{ - CMtCoderThread *t = (CMtCoderThread *)pp; - for (;;) - { - if (Event_Wait(&t->startEvent) != 0) - return (THREAD_FUNC_RET_TYPE)SZ_ERROR_THREAD; - if (t->stop) - return 0; - { - SRes res = ThreadFunc2(t); - CMtCoder *mtc = t->mtCoder; - if (res != SZ_OK) - { - MtProgress_SetError(&mtc->mtProgress, res); - } - - #ifndef MTCODER__USE_WRITE_THREAD - { - unsigned numFinished = (unsigned)InterlockedIncrement(&mtc->numFinishedThreads); - if (numFinished == mtc->numStartedThreads) - if (Event_Set(&mtc->finishedEvent) != 0) - return (THREAD_FUNC_RET_TYPE)SZ_ERROR_THREAD; - } - #endif - } - } -} - - - -void MtCoder_Construct(CMtCoder *p) -{ - unsigned i; - - p->blockSize = 0; - p->numThreadsMax = 0; - p->expectedDataSize = (UInt64)(Int64)-1; - - p->inStream = NULL; - p->inData = NULL; - p->inDataSize = 0; - - p->progress = NULL; - p->allocBig = NULL; - - p->mtCallback = NULL; - p->mtCallbackObject = NULL; - - p->allocatedBufsSize = 0; - - Event_Construct(&p->readEvent); - Semaphore_Construct(&p->blocksSemaphore); - - for (i = 0; i < MTCODER__THREADS_MAX; i++) - { - CMtCoderThread *t = &p->threads[i]; - t->mtCoder = p; - t->index = i; - t->inBuf = NULL; - t->stop = False; - Event_Construct(&t->startEvent); - Thread_Construct(&t->thread); - } - - #ifdef MTCODER__USE_WRITE_THREAD - for (i = 0; i < MTCODER__BLOCKS_MAX; i++) - Event_Construct(&p->writeEvents[i]); - #else - Event_Construct(&p->finishedEvent); - #endif - - CriticalSection_Init(&p->cs); - CriticalSection_Init(&p->mtProgress.cs); -} - - - - -static void MtCoder_Free(CMtCoder *p) -{ - unsigned i; - - /* - p->stopReading = True; - if (Event_IsCreated(&p->readEvent)) - Event_Set(&p->readEvent); - */ - - for (i = 0; i < MTCODER__THREADS_MAX; i++) - MtCoderThread_Destruct(&p->threads[i]); - - Event_Close(&p->readEvent); - Semaphore_Close(&p->blocksSemaphore); - - #ifdef MTCODER__USE_WRITE_THREAD - for (i = 0; i < MTCODER__BLOCKS_MAX; i++) - Event_Close(&p->writeEvents[i]); - #else - Event_Close(&p->finishedEvent); - #endif -} - - -void MtCoder_Destruct(CMtCoder *p) -{ - MtCoder_Free(p); - - CriticalSection_Delete(&p->cs); - CriticalSection_Delete(&p->mtProgress.cs); -} - - -SRes MtCoder_Code(CMtCoder *p) -{ - unsigned numThreads = p->numThreadsMax; - unsigned numBlocksMax; - unsigned i; - SRes res = SZ_OK; - - if (numThreads > MTCODER__THREADS_MAX) - numThreads = MTCODER__THREADS_MAX; - numBlocksMax = MTCODER__GET_NUM_BLOCKS_FROM_THREADS(numThreads); - - if (p->blockSize < ((UInt32)1 << 26)) numBlocksMax++; - if (p->blockSize < ((UInt32)1 << 24)) numBlocksMax++; - if (p->blockSize < ((UInt32)1 << 22)) numBlocksMax++; - - if (numBlocksMax > MTCODER__BLOCKS_MAX) - numBlocksMax = MTCODER__BLOCKS_MAX; - - if (p->blockSize != p->allocatedBufsSize) - { - for (i = 0; i < MTCODER__THREADS_MAX; i++) - { - CMtCoderThread *t = &p->threads[i]; - if (t->inBuf) - { - ISzAlloc_Free(p->allocBig, t->inBuf); - t->inBuf = NULL; - } - } - p->allocatedBufsSize = p->blockSize; - } - - p->readRes = SZ_OK; - - MtProgress_Init(&p->mtProgress, p->progress); - - #ifdef MTCODER__USE_WRITE_THREAD - for (i = 0; i < numBlocksMax; i++) - { - RINOK_THREAD(ArEvent_OptCreate_And_Reset(&p->writeEvents[i])); - } - #else - RINOK_THREAD(ArEvent_OptCreate_And_Reset(&p->finishedEvent)); - #endif - - { - RINOK_THREAD(ArEvent_OptCreate_And_Reset(&p->readEvent)); - RINOK_THREAD(Semaphore_OptCreateInit(&p->blocksSemaphore, numBlocksMax, numBlocksMax)); - } - - for (i = 0; i < MTCODER__BLOCKS_MAX - 1; i++) - p->freeBlockList[i] = i + 1; - p->freeBlockList[MTCODER__BLOCKS_MAX - 1] = (unsigned)(int)-1; - p->freeBlockHead = 0; - - p->readProcessed = 0; - p->blockIndex = 0; - p->numBlocksMax = numBlocksMax; - p->stopReading = False; - - #ifndef MTCODER__USE_WRITE_THREAD - p->writeIndex = 0; - p->writeRes = SZ_OK; - for (i = 0; i < MTCODER__BLOCKS_MAX; i++) - p->ReadyBlocks[i] = False; - p->numFinishedThreads = 0; - #endif - - p->numStartedThreadsLimit = numThreads; - p->numStartedThreads = 0; - - // for (i = 0; i < numThreads; i++) - { - CMtCoderThread *nextThread = &p->threads[p->numStartedThreads++]; - RINOK(MtCoderThread_CreateAndStart(nextThread)); - } - - RINOK_THREAD(Event_Set(&p->readEvent)) - - #ifdef MTCODER__USE_WRITE_THREAD - { - unsigned bi = 0; - - for (;; bi++) - { - if (bi >= numBlocksMax) - bi = 0; - - RINOK_THREAD(Event_Wait(&p->writeEvents[bi])) - - { - const CMtCoderBlock *block = &p->blocks[bi]; - unsigned bufIndex = block->bufIndex; - BoolInt finished = block->finished; - if (res == SZ_OK && block->res != SZ_OK) - res = block->res; - - if (bufIndex != (unsigned)(int)-1) - { - if (res == SZ_OK) - { - res = p->mtCallback->Write(p->mtCallbackObject, bufIndex); - if (res != SZ_OK) - MtProgress_SetError(&p->mtProgress, res); - } - - CriticalSection_Enter(&p->cs); - { - p->freeBlockList[bufIndex] = p->freeBlockHead; - p->freeBlockHead = bufIndex; - } - CriticalSection_Leave(&p->cs); - } - - RINOK_THREAD(Semaphore_Release1(&p->blocksSemaphore)) - - if (finished) - break; - } - } - } - #else - { - WRes wres = Event_Wait(&p->finishedEvent); - res = MY_SRes_HRESULT_FROM_WRes(wres); - } - #endif - - if (res == SZ_OK) - res = p->readRes; - - if (res == SZ_OK) - res = p->mtProgress.res; - - #ifndef MTCODER__USE_WRITE_THREAD - if (res == SZ_OK) - res = p->writeRes; - #endif - - if (res != SZ_OK) - MtCoder_Free(p); - return res; -} - -#endif +/* MtCoder.c -- Multi-thread Coder +2021-12-21 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "MtCoder.h" + +#ifndef _7ZIP_ST + +static SRes MtProgressThunk_Progress(const ICompressProgress *pp, UInt64 inSize, UInt64 outSize) +{ + CMtProgressThunk *thunk = CONTAINER_FROM_VTBL(pp, CMtProgressThunk, vt); + UInt64 inSize2 = 0; + UInt64 outSize2 = 0; + if (inSize != (UInt64)(Int64)-1) + { + inSize2 = inSize - thunk->inSize; + thunk->inSize = inSize; + } + if (outSize != (UInt64)(Int64)-1) + { + outSize2 = outSize - thunk->outSize; + thunk->outSize = outSize; + } + return MtProgress_ProgressAdd(thunk->mtProgress, inSize2, outSize2); +} + + +void MtProgressThunk_CreateVTable(CMtProgressThunk *p) +{ + p->vt.Progress = MtProgressThunk_Progress; +} + + + +#define RINOK_THREAD(x) { if ((x) != 0) return SZ_ERROR_THREAD; } + + +static WRes ArEvent_OptCreate_And_Reset(CEvent *p) +{ + if (Event_IsCreated(p)) + return Event_Reset(p); + return AutoResetEvent_CreateNotSignaled(p); +} + + +static THREAD_FUNC_DECL ThreadFunc(void *pp); + + +static SRes MtCoderThread_CreateAndStart(CMtCoderThread *t) +{ + WRes wres = ArEvent_OptCreate_And_Reset(&t->startEvent); + if (wres == 0) + { + t->stop = False; + if (!Thread_WasCreated(&t->thread)) + wres = Thread_Create(&t->thread, ThreadFunc, t); + if (wres == 0) + wres = Event_Set(&t->startEvent); + } + if (wres == 0) + return SZ_OK; + return MY_SRes_HRESULT_FROM_WRes(wres); +} + + +static void MtCoderThread_Destruct(CMtCoderThread *t) +{ + if (Thread_WasCreated(&t->thread)) + { + t->stop = 1; + Event_Set(&t->startEvent); + Thread_Wait_Close(&t->thread); + } + + Event_Close(&t->startEvent); + + if (t->inBuf) + { + ISzAlloc_Free(t->mtCoder->allocBig, t->inBuf); + t->inBuf = NULL; + } +} + + + +static SRes FullRead(ISeqInStream *stream, Byte *data, size_t *processedSize) +{ + size_t size = *processedSize; + *processedSize = 0; + while (size != 0) + { + size_t cur = size; + SRes res = ISeqInStream_Read(stream, data, &cur); + *processedSize += cur; + data += cur; + size -= cur; + RINOK(res); + if (cur == 0) + return SZ_OK; + } + return SZ_OK; +} + + +/* + ThreadFunc2() returns: + SZ_OK - in all normal cases (even for stream error or memory allocation error) + SZ_ERROR_THREAD - in case of failure in system synch function +*/ + +static SRes ThreadFunc2(CMtCoderThread *t) +{ + CMtCoder *mtc = t->mtCoder; + + for (;;) + { + unsigned bi; + SRes res; + SRes res2; + BoolInt finished; + unsigned bufIndex; + size_t size; + const Byte *inData; + UInt64 readProcessed = 0; + + RINOK_THREAD(Event_Wait(&mtc->readEvent)) + + /* after Event_Wait(&mtc->readEvent) we must call Event_Set(&mtc->readEvent) in any case to unlock another threads */ + + if (mtc->stopReading) + { + return Event_Set(&mtc->readEvent) == 0 ? SZ_OK : SZ_ERROR_THREAD; + } + + res = MtProgress_GetError(&mtc->mtProgress); + + size = 0; + inData = NULL; + finished = True; + + if (res == SZ_OK) + { + size = mtc->blockSize; + if (mtc->inStream) + { + if (!t->inBuf) + { + t->inBuf = (Byte *)ISzAlloc_Alloc(mtc->allocBig, mtc->blockSize); + if (!t->inBuf) + res = SZ_ERROR_MEM; + } + if (res == SZ_OK) + { + res = FullRead(mtc->inStream, t->inBuf, &size); + readProcessed = mtc->readProcessed + size; + mtc->readProcessed = readProcessed; + } + if (res != SZ_OK) + { + mtc->readRes = res; + /* after reading error - we can stop encoding of previous blocks */ + MtProgress_SetError(&mtc->mtProgress, res); + } + else + finished = (size != mtc->blockSize); + } + else + { + size_t rem; + readProcessed = mtc->readProcessed; + rem = mtc->inDataSize - (size_t)readProcessed; + if (size > rem) + size = rem; + inData = mtc->inData + (size_t)readProcessed; + readProcessed += size; + mtc->readProcessed = readProcessed; + finished = (mtc->inDataSize == (size_t)readProcessed); + } + } + + /* we must get some block from blocksSemaphore before Event_Set(&mtc->readEvent) */ + + res2 = SZ_OK; + + if (Semaphore_Wait(&mtc->blocksSemaphore) != 0) + { + res2 = SZ_ERROR_THREAD; + if (res == SZ_OK) + { + res = res2; + // MtProgress_SetError(&mtc->mtProgress, res); + } + } + + bi = mtc->blockIndex; + + if (++mtc->blockIndex >= mtc->numBlocksMax) + mtc->blockIndex = 0; + + bufIndex = (unsigned)(int)-1; + + if (res == SZ_OK) + res = MtProgress_GetError(&mtc->mtProgress); + + if (res != SZ_OK) + finished = True; + + if (!finished) + { + if (mtc->numStartedThreads < mtc->numStartedThreadsLimit + && mtc->expectedDataSize != readProcessed) + { + res = MtCoderThread_CreateAndStart(&mtc->threads[mtc->numStartedThreads]); + if (res == SZ_OK) + mtc->numStartedThreads++; + else + { + MtProgress_SetError(&mtc->mtProgress, res); + finished = True; + } + } + } + + if (finished) + mtc->stopReading = True; + + RINOK_THREAD(Event_Set(&mtc->readEvent)) + + if (res2 != SZ_OK) + return res2; + + if (res == SZ_OK) + { + CriticalSection_Enter(&mtc->cs); + bufIndex = mtc->freeBlockHead; + mtc->freeBlockHead = mtc->freeBlockList[bufIndex]; + CriticalSection_Leave(&mtc->cs); + + res = mtc->mtCallback->Code(mtc->mtCallbackObject, t->index, bufIndex, + mtc->inStream ? t->inBuf : inData, size, finished); + + // MtProgress_Reinit(&mtc->mtProgress, t->index); + + if (res != SZ_OK) + MtProgress_SetError(&mtc->mtProgress, res); + } + + { + CMtCoderBlock *block = &mtc->blocks[bi]; + block->res = res; + block->bufIndex = bufIndex; + block->finished = finished; + } + + #ifdef MTCODER__USE_WRITE_THREAD + RINOK_THREAD(Event_Set(&mtc->writeEvents[bi])) + #else + { + unsigned wi; + { + CriticalSection_Enter(&mtc->cs); + wi = mtc->writeIndex; + if (wi == bi) + mtc->writeIndex = (unsigned)(int)-1; + else + mtc->ReadyBlocks[bi] = True; + CriticalSection_Leave(&mtc->cs); + } + + if (wi != bi) + { + if (res != SZ_OK || finished) + return 0; + continue; + } + + if (mtc->writeRes != SZ_OK) + res = mtc->writeRes; + + for (;;) + { + if (res == SZ_OK && bufIndex != (unsigned)(int)-1) + { + res = mtc->mtCallback->Write(mtc->mtCallbackObject, bufIndex); + if (res != SZ_OK) + { + mtc->writeRes = res; + MtProgress_SetError(&mtc->mtProgress, res); + } + } + + if (++wi >= mtc->numBlocksMax) + wi = 0; + { + BoolInt isReady; + + CriticalSection_Enter(&mtc->cs); + + if (bufIndex != (unsigned)(int)-1) + { + mtc->freeBlockList[bufIndex] = mtc->freeBlockHead; + mtc->freeBlockHead = bufIndex; + } + + isReady = mtc->ReadyBlocks[wi]; + + if (isReady) + mtc->ReadyBlocks[wi] = False; + else + mtc->writeIndex = wi; + + CriticalSection_Leave(&mtc->cs); + + RINOK_THREAD(Semaphore_Release1(&mtc->blocksSemaphore)) + + if (!isReady) + break; + } + + { + CMtCoderBlock *block = &mtc->blocks[wi]; + if (res == SZ_OK && block->res != SZ_OK) + res = block->res; + bufIndex = block->bufIndex; + finished = block->finished; + } + } + } + #endif + + if (finished || res != SZ_OK) + return 0; + } +} + + +static THREAD_FUNC_DECL ThreadFunc(void *pp) +{ + CMtCoderThread *t = (CMtCoderThread *)pp; + for (;;) + { + if (Event_Wait(&t->startEvent) != 0) + return (THREAD_FUNC_RET_TYPE)SZ_ERROR_THREAD; + if (t->stop) + return 0; + { + SRes res = ThreadFunc2(t); + CMtCoder *mtc = t->mtCoder; + if (res != SZ_OK) + { + MtProgress_SetError(&mtc->mtProgress, res); + } + + #ifndef MTCODER__USE_WRITE_THREAD + { + unsigned numFinished = (unsigned)InterlockedIncrement(&mtc->numFinishedThreads); + if (numFinished == mtc->numStartedThreads) + if (Event_Set(&mtc->finishedEvent) != 0) + return (THREAD_FUNC_RET_TYPE)SZ_ERROR_THREAD; + } + #endif + } + } +} + + + +void MtCoder_Construct(CMtCoder *p) +{ + unsigned i; + + p->blockSize = 0; + p->numThreadsMax = 0; + p->expectedDataSize = (UInt64)(Int64)-1; + + p->inStream = NULL; + p->inData = NULL; + p->inDataSize = 0; + + p->progress = NULL; + p->allocBig = NULL; + + p->mtCallback = NULL; + p->mtCallbackObject = NULL; + + p->allocatedBufsSize = 0; + + Event_Construct(&p->readEvent); + Semaphore_Construct(&p->blocksSemaphore); + + for (i = 0; i < MTCODER__THREADS_MAX; i++) + { + CMtCoderThread *t = &p->threads[i]; + t->mtCoder = p; + t->index = i; + t->inBuf = NULL; + t->stop = False; + Event_Construct(&t->startEvent); + Thread_Construct(&t->thread); + } + + #ifdef MTCODER__USE_WRITE_THREAD + for (i = 0; i < MTCODER__BLOCKS_MAX; i++) + Event_Construct(&p->writeEvents[i]); + #else + Event_Construct(&p->finishedEvent); + #endif + + CriticalSection_Init(&p->cs); + CriticalSection_Init(&p->mtProgress.cs); +} + + + + +static void MtCoder_Free(CMtCoder *p) +{ + unsigned i; + + /* + p->stopReading = True; + if (Event_IsCreated(&p->readEvent)) + Event_Set(&p->readEvent); + */ + + for (i = 0; i < MTCODER__THREADS_MAX; i++) + MtCoderThread_Destruct(&p->threads[i]); + + Event_Close(&p->readEvent); + Semaphore_Close(&p->blocksSemaphore); + + #ifdef MTCODER__USE_WRITE_THREAD + for (i = 0; i < MTCODER__BLOCKS_MAX; i++) + Event_Close(&p->writeEvents[i]); + #else + Event_Close(&p->finishedEvent); + #endif +} + + +void MtCoder_Destruct(CMtCoder *p) +{ + MtCoder_Free(p); + + CriticalSection_Delete(&p->cs); + CriticalSection_Delete(&p->mtProgress.cs); +} + + +SRes MtCoder_Code(CMtCoder *p) +{ + unsigned numThreads = p->numThreadsMax; + unsigned numBlocksMax; + unsigned i; + SRes res = SZ_OK; + + if (numThreads > MTCODER__THREADS_MAX) + numThreads = MTCODER__THREADS_MAX; + numBlocksMax = MTCODER__GET_NUM_BLOCKS_FROM_THREADS(numThreads); + + if (p->blockSize < ((UInt32)1 << 26)) numBlocksMax++; + if (p->blockSize < ((UInt32)1 << 24)) numBlocksMax++; + if (p->blockSize < ((UInt32)1 << 22)) numBlocksMax++; + + if (numBlocksMax > MTCODER__BLOCKS_MAX) + numBlocksMax = MTCODER__BLOCKS_MAX; + + if (p->blockSize != p->allocatedBufsSize) + { + for (i = 0; i < MTCODER__THREADS_MAX; i++) + { + CMtCoderThread *t = &p->threads[i]; + if (t->inBuf) + { + ISzAlloc_Free(p->allocBig, t->inBuf); + t->inBuf = NULL; + } + } + p->allocatedBufsSize = p->blockSize; + } + + p->readRes = SZ_OK; + + MtProgress_Init(&p->mtProgress, p->progress); + + #ifdef MTCODER__USE_WRITE_THREAD + for (i = 0; i < numBlocksMax; i++) + { + RINOK_THREAD(ArEvent_OptCreate_And_Reset(&p->writeEvents[i])); + } + #else + RINOK_THREAD(ArEvent_OptCreate_And_Reset(&p->finishedEvent)); + #endif + + { + RINOK_THREAD(ArEvent_OptCreate_And_Reset(&p->readEvent)); + RINOK_THREAD(Semaphore_OptCreateInit(&p->blocksSemaphore, numBlocksMax, numBlocksMax)); + } + + for (i = 0; i < MTCODER__BLOCKS_MAX - 1; i++) + p->freeBlockList[i] = i + 1; + p->freeBlockList[MTCODER__BLOCKS_MAX - 1] = (unsigned)(int)-1; + p->freeBlockHead = 0; + + p->readProcessed = 0; + p->blockIndex = 0; + p->numBlocksMax = numBlocksMax; + p->stopReading = False; + + #ifndef MTCODER__USE_WRITE_THREAD + p->writeIndex = 0; + p->writeRes = SZ_OK; + for (i = 0; i < MTCODER__BLOCKS_MAX; i++) + p->ReadyBlocks[i] = False; + p->numFinishedThreads = 0; + #endif + + p->numStartedThreadsLimit = numThreads; + p->numStartedThreads = 0; + + // for (i = 0; i < numThreads; i++) + { + CMtCoderThread *nextThread = &p->threads[p->numStartedThreads++]; + RINOK(MtCoderThread_CreateAndStart(nextThread)); + } + + RINOK_THREAD(Event_Set(&p->readEvent)) + + #ifdef MTCODER__USE_WRITE_THREAD + { + unsigned bi = 0; + + for (;; bi++) + { + if (bi >= numBlocksMax) + bi = 0; + + RINOK_THREAD(Event_Wait(&p->writeEvents[bi])) + + { + const CMtCoderBlock *block = &p->blocks[bi]; + unsigned bufIndex = block->bufIndex; + BoolInt finished = block->finished; + if (res == SZ_OK && block->res != SZ_OK) + res = block->res; + + if (bufIndex != (unsigned)(int)-1) + { + if (res == SZ_OK) + { + res = p->mtCallback->Write(p->mtCallbackObject, bufIndex); + if (res != SZ_OK) + MtProgress_SetError(&p->mtProgress, res); + } + + CriticalSection_Enter(&p->cs); + { + p->freeBlockList[bufIndex] = p->freeBlockHead; + p->freeBlockHead = bufIndex; + } + CriticalSection_Leave(&p->cs); + } + + RINOK_THREAD(Semaphore_Release1(&p->blocksSemaphore)) + + if (finished) + break; + } + } + } + #else + { + WRes wres = Event_Wait(&p->finishedEvent); + res = MY_SRes_HRESULT_FROM_WRes(wres); + } + #endif + + if (res == SZ_OK) + res = p->readRes; + + if (res == SZ_OK) + res = p->mtProgress.res; + + #ifndef MTCODER__USE_WRITE_THREAD + if (res == SZ_OK) + res = p->writeRes; + #endif + + if (res != SZ_OK) + MtCoder_Free(p); + return res; +} + +#endif diff --git a/deps/LZMA-SDK/C/MtCoder.h b/deps/LZMA-SDK/C/MtCoder.h index 603329d36..5a5f4d11b 100644 --- a/deps/LZMA-SDK/C/MtCoder.h +++ b/deps/LZMA-SDK/C/MtCoder.h @@ -1,141 +1,141 @@ -/* MtCoder.h -- Multi-thread Coder -2018-07-04 : Igor Pavlov : Public domain */ - -#ifndef __MT_CODER_H -#define __MT_CODER_H - -#include "MtDec.h" - -EXTERN_C_BEGIN - -/* - if ( defined MTCODER__USE_WRITE_THREAD) : main thread writes all data blocks to output stream - if (not defined MTCODER__USE_WRITE_THREAD) : any coder thread can write data blocks to output stream -*/ -/* #define MTCODER__USE_WRITE_THREAD */ - -#ifndef _7ZIP_ST - #define MTCODER__GET_NUM_BLOCKS_FROM_THREADS(numThreads) ((numThreads) + (numThreads) / 8 + 1) - #define MTCODER__THREADS_MAX 64 - #define MTCODER__BLOCKS_MAX (MTCODER__GET_NUM_BLOCKS_FROM_THREADS(MTCODER__THREADS_MAX) + 3) -#else - #define MTCODER__THREADS_MAX 1 - #define MTCODER__BLOCKS_MAX 1 -#endif - - -#ifndef _7ZIP_ST - - -typedef struct -{ - ICompressProgress vt; - CMtProgress *mtProgress; - UInt64 inSize; - UInt64 outSize; -} CMtProgressThunk; - -void MtProgressThunk_CreateVTable(CMtProgressThunk *p); - -#define MtProgressThunk_Init(p) { (p)->inSize = 0; (p)->outSize = 0; } - - -struct _CMtCoder; - - -typedef struct -{ - struct _CMtCoder *mtCoder; - unsigned index; - int stop; - Byte *inBuf; - - CAutoResetEvent startEvent; - CThread thread; -} CMtCoderThread; - - -typedef struct -{ - SRes (*Code)(void *p, unsigned coderIndex, unsigned outBufIndex, - const Byte *src, size_t srcSize, int finished); - SRes (*Write)(void *p, unsigned outBufIndex); -} IMtCoderCallback2; - - -typedef struct -{ - SRes res; - unsigned bufIndex; - BoolInt finished; -} CMtCoderBlock; - - -typedef struct _CMtCoder -{ - /* input variables */ - - size_t blockSize; /* size of input block */ - unsigned numThreadsMax; - UInt64 expectedDataSize; - - ISeqInStream *inStream; - const Byte *inData; - size_t inDataSize; - - ICompressProgress *progress; - ISzAllocPtr allocBig; - - IMtCoderCallback2 *mtCallback; - void *mtCallbackObject; - - - /* internal variables */ - - size_t allocatedBufsSize; - - CAutoResetEvent readEvent; - CSemaphore blocksSemaphore; - - BoolInt stopReading; - SRes readRes; - - #ifdef MTCODER__USE_WRITE_THREAD - CAutoResetEvent writeEvents[MTCODER__BLOCKS_MAX]; - #else - CAutoResetEvent finishedEvent; - SRes writeRes; - unsigned writeIndex; - Byte ReadyBlocks[MTCODER__BLOCKS_MAX]; - LONG numFinishedThreads; - #endif - - unsigned numStartedThreadsLimit; - unsigned numStartedThreads; - - unsigned numBlocksMax; - unsigned blockIndex; - UInt64 readProcessed; - - CCriticalSection cs; - - unsigned freeBlockHead; - unsigned freeBlockList[MTCODER__BLOCKS_MAX]; - - CMtProgress mtProgress; - CMtCoderBlock blocks[MTCODER__BLOCKS_MAX]; - CMtCoderThread threads[MTCODER__THREADS_MAX]; -} CMtCoder; - - -void MtCoder_Construct(CMtCoder *p); -void MtCoder_Destruct(CMtCoder *p); -SRes MtCoder_Code(CMtCoder *p); - - -#endif - - -EXTERN_C_END - -#endif +/* MtCoder.h -- Multi-thread Coder +2018-07-04 : Igor Pavlov : Public domain */ + +#ifndef __MT_CODER_H +#define __MT_CODER_H + +#include "MtDec.h" + +EXTERN_C_BEGIN + +/* + if ( defined MTCODER__USE_WRITE_THREAD) : main thread writes all data blocks to output stream + if (not defined MTCODER__USE_WRITE_THREAD) : any coder thread can write data blocks to output stream +*/ +/* #define MTCODER__USE_WRITE_THREAD */ + +#ifndef _7ZIP_ST + #define MTCODER__GET_NUM_BLOCKS_FROM_THREADS(numThreads) ((numThreads) + (numThreads) / 8 + 1) + #define MTCODER__THREADS_MAX 64 + #define MTCODER__BLOCKS_MAX (MTCODER__GET_NUM_BLOCKS_FROM_THREADS(MTCODER__THREADS_MAX) + 3) +#else + #define MTCODER__THREADS_MAX 1 + #define MTCODER__BLOCKS_MAX 1 +#endif + + +#ifndef _7ZIP_ST + + +typedef struct +{ + ICompressProgress vt; + CMtProgress *mtProgress; + UInt64 inSize; + UInt64 outSize; +} CMtProgressThunk; + +void MtProgressThunk_CreateVTable(CMtProgressThunk *p); + +#define MtProgressThunk_Init(p) { (p)->inSize = 0; (p)->outSize = 0; } + + +struct _CMtCoder; + + +typedef struct +{ + struct _CMtCoder *mtCoder; + unsigned index; + int stop; + Byte *inBuf; + + CAutoResetEvent startEvent; + CThread thread; +} CMtCoderThread; + + +typedef struct +{ + SRes (*Code)(void *p, unsigned coderIndex, unsigned outBufIndex, + const Byte *src, size_t srcSize, int finished); + SRes (*Write)(void *p, unsigned outBufIndex); +} IMtCoderCallback2; + + +typedef struct +{ + SRes res; + unsigned bufIndex; + BoolInt finished; +} CMtCoderBlock; + + +typedef struct _CMtCoder +{ + /* input variables */ + + size_t blockSize; /* size of input block */ + unsigned numThreadsMax; + UInt64 expectedDataSize; + + ISeqInStream *inStream; + const Byte *inData; + size_t inDataSize; + + ICompressProgress *progress; + ISzAllocPtr allocBig; + + IMtCoderCallback2 *mtCallback; + void *mtCallbackObject; + + + /* internal variables */ + + size_t allocatedBufsSize; + + CAutoResetEvent readEvent; + CSemaphore blocksSemaphore; + + BoolInt stopReading; + SRes readRes; + + #ifdef MTCODER__USE_WRITE_THREAD + CAutoResetEvent writeEvents[MTCODER__BLOCKS_MAX]; + #else + CAutoResetEvent finishedEvent; + SRes writeRes; + unsigned writeIndex; + Byte ReadyBlocks[MTCODER__BLOCKS_MAX]; + LONG numFinishedThreads; + #endif + + unsigned numStartedThreadsLimit; + unsigned numStartedThreads; + + unsigned numBlocksMax; + unsigned blockIndex; + UInt64 readProcessed; + + CCriticalSection cs; + + unsigned freeBlockHead; + unsigned freeBlockList[MTCODER__BLOCKS_MAX]; + + CMtProgress mtProgress; + CMtCoderBlock blocks[MTCODER__BLOCKS_MAX]; + CMtCoderThread threads[MTCODER__THREADS_MAX]; +} CMtCoder; + + +void MtCoder_Construct(CMtCoder *p); +void MtCoder_Destruct(CMtCoder *p); +SRes MtCoder_Code(CMtCoder *p); + + +#endif + + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/MtDec.c b/deps/LZMA-SDK/C/MtDec.c index 854087b92..45a671396 100644 --- a/deps/LZMA-SDK/C/MtDec.c +++ b/deps/LZMA-SDK/C/MtDec.c @@ -1,1139 +1,1139 @@ -/* MtDec.c -- Multi-thread Decoder -2021-12-21 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -// #define SHOW_DEBUG_INFO - -// #include -#include - -#ifdef SHOW_DEBUG_INFO -#include -#endif - -#include "MtDec.h" - -#ifndef _7ZIP_ST - -#ifdef SHOW_DEBUG_INFO -#define PRF(x) x -#else -#define PRF(x) -#endif - -#define PRF_STR_INT(s, d) PRF(printf("\n" s " %d\n", (unsigned)d)) - -void MtProgress_Init(CMtProgress *p, ICompressProgress *progress) -{ - p->progress = progress; - p->res = SZ_OK; - p->totalInSize = 0; - p->totalOutSize = 0; -} - - -SRes MtProgress_Progress_ST(CMtProgress *p) -{ - if (p->res == SZ_OK && p->progress) - if (ICompressProgress_Progress(p->progress, p->totalInSize, p->totalOutSize) != SZ_OK) - p->res = SZ_ERROR_PROGRESS; - return p->res; -} - - -SRes MtProgress_ProgressAdd(CMtProgress *p, UInt64 inSize, UInt64 outSize) -{ - SRes res; - CriticalSection_Enter(&p->cs); - - p->totalInSize += inSize; - p->totalOutSize += outSize; - if (p->res == SZ_OK && p->progress) - if (ICompressProgress_Progress(p->progress, p->totalInSize, p->totalOutSize) != SZ_OK) - p->res = SZ_ERROR_PROGRESS; - res = p->res; - - CriticalSection_Leave(&p->cs); - return res; -} - - -SRes MtProgress_GetError(CMtProgress *p) -{ - SRes res; - CriticalSection_Enter(&p->cs); - res = p->res; - CriticalSection_Leave(&p->cs); - return res; -} - - -void MtProgress_SetError(CMtProgress *p, SRes res) -{ - CriticalSection_Enter(&p->cs); - if (p->res == SZ_OK) - p->res = res; - CriticalSection_Leave(&p->cs); -} - - -#define RINOK_THREAD(x) RINOK_WRes(x) - - -static WRes ArEvent_OptCreate_And_Reset(CEvent *p) -{ - if (Event_IsCreated(p)) - return Event_Reset(p); - return AutoResetEvent_CreateNotSignaled(p); -} - - -struct __CMtDecBufLink -{ - struct __CMtDecBufLink *next; - void *pad[3]; -}; - -typedef struct __CMtDecBufLink CMtDecBufLink; - -#define MTDEC__LINK_DATA_OFFSET sizeof(CMtDecBufLink) -#define MTDEC__DATA_PTR_FROM_LINK(link) ((Byte *)(link) + MTDEC__LINK_DATA_OFFSET) - - - -static THREAD_FUNC_DECL ThreadFunc(void *pp); - - -static WRes MtDecThread_CreateEvents(CMtDecThread *t) -{ - WRes wres = ArEvent_OptCreate_And_Reset(&t->canWrite); - if (wres == 0) - { - wres = ArEvent_OptCreate_And_Reset(&t->canRead); - if (wres == 0) - return SZ_OK; - } - return wres; -} - - -static SRes MtDecThread_CreateAndStart(CMtDecThread *t) -{ - WRes wres = MtDecThread_CreateEvents(t); - // wres = 17; // for test - if (wres == 0) - { - if (Thread_WasCreated(&t->thread)) - return SZ_OK; - wres = Thread_Create(&t->thread, ThreadFunc, t); - if (wres == 0) - return SZ_OK; - } - return MY_SRes_HRESULT_FROM_WRes(wres); -} - - -void MtDecThread_FreeInBufs(CMtDecThread *t) -{ - if (t->inBuf) - { - void *link = t->inBuf; - t->inBuf = NULL; - do - { - void *next = ((CMtDecBufLink *)link)->next; - ISzAlloc_Free(t->mtDec->alloc, link); - link = next; - } - while (link); - } -} - - -static void MtDecThread_CloseThread(CMtDecThread *t) -{ - if (Thread_WasCreated(&t->thread)) - { - Event_Set(&t->canWrite); /* we can disable it. There are no threads waiting canWrite in normal cases */ - Event_Set(&t->canRead); - Thread_Wait_Close(&t->thread); - } - - Event_Close(&t->canRead); - Event_Close(&t->canWrite); -} - -static void MtDec_CloseThreads(CMtDec *p) -{ - unsigned i; - for (i = 0; i < MTDEC__THREADS_MAX; i++) - MtDecThread_CloseThread(&p->threads[i]); -} - -static void MtDecThread_Destruct(CMtDecThread *t) -{ - MtDecThread_CloseThread(t); - MtDecThread_FreeInBufs(t); -} - - - -static SRes FullRead(ISeqInStream *stream, Byte *data, size_t *processedSize) -{ - size_t size = *processedSize; - *processedSize = 0; - while (size != 0) - { - size_t cur = size; - SRes res = ISeqInStream_Read(stream, data, &cur); - *processedSize += cur; - data += cur; - size -= cur; - RINOK(res); - if (cur == 0) - return SZ_OK; - } - return SZ_OK; -} - - -static SRes MtDec_GetError_Spec(CMtDec *p, UInt64 interruptIndex, BoolInt *wasInterrupted) -{ - SRes res; - CriticalSection_Enter(&p->mtProgress.cs); - *wasInterrupted = (p->needInterrupt && interruptIndex > p->interruptIndex); - res = p->mtProgress.res; - CriticalSection_Leave(&p->mtProgress.cs); - return res; -} - -static SRes MtDec_Progress_GetError_Spec(CMtDec *p, UInt64 inSize, UInt64 outSize, UInt64 interruptIndex, BoolInt *wasInterrupted) -{ - SRes res; - CriticalSection_Enter(&p->mtProgress.cs); - - p->mtProgress.totalInSize += inSize; - p->mtProgress.totalOutSize += outSize; - if (p->mtProgress.res == SZ_OK && p->mtProgress.progress) - if (ICompressProgress_Progress(p->mtProgress.progress, p->mtProgress.totalInSize, p->mtProgress.totalOutSize) != SZ_OK) - p->mtProgress.res = SZ_ERROR_PROGRESS; - - *wasInterrupted = (p->needInterrupt && interruptIndex > p->interruptIndex); - res = p->mtProgress.res; - - CriticalSection_Leave(&p->mtProgress.cs); - - return res; -} - -static void MtDec_Interrupt(CMtDec *p, UInt64 interruptIndex) -{ - CriticalSection_Enter(&p->mtProgress.cs); - if (!p->needInterrupt || interruptIndex < p->interruptIndex) - { - p->interruptIndex = interruptIndex; - p->needInterrupt = True; - } - CriticalSection_Leave(&p->mtProgress.cs); -} - -Byte *MtDec_GetCrossBuff(CMtDec *p) -{ - Byte *cr = p->crossBlock; - if (!cr) - { - cr = (Byte *)ISzAlloc_Alloc(p->alloc, MTDEC__LINK_DATA_OFFSET + p->inBufSize); - if (!cr) - return NULL; - p->crossBlock = cr; - } - return MTDEC__DATA_PTR_FROM_LINK(cr); -} - - -/* - ThreadFunc2() returns: - 0 - in all normal cases (even for stream error or memory allocation error) - (!= 0) - WRes error return by system threading function -*/ - -// #define MTDEC_ProgessStep (1 << 22) -#define MTDEC_ProgessStep (1 << 0) - -static WRes ThreadFunc2(CMtDecThread *t) -{ - CMtDec *p = t->mtDec; - - PRF_STR_INT("ThreadFunc2", t->index); - - // SetThreadAffinityMask(GetCurrentThread(), 1 << t->index); - - for (;;) - { - SRes res, codeRes; - BoolInt wasInterrupted, isAllocError, overflow, finish; - SRes threadingErrorSRes; - BoolInt needCode, needWrite, needContinue; - - size_t inDataSize_Start; - UInt64 inDataSize; - // UInt64 inDataSize_Full; - - UInt64 blockIndex; - - UInt64 inPrev = 0; - UInt64 outPrev = 0; - UInt64 inCodePos; - UInt64 outCodePos; - - Byte *afterEndData = NULL; - size_t afterEndData_Size = 0; - BoolInt afterEndData_IsCross = False; - - BoolInt canCreateNewThread = False; - // CMtDecCallbackInfo parse; - CMtDecThread *nextThread; - - PRF_STR_INT("=============== Event_Wait(&t->canRead)", t->index); - - RINOK_THREAD(Event_Wait(&t->canRead)); - if (p->exitThread) - return 0; - - PRF_STR_INT("after Event_Wait(&t->canRead)", t->index); - - // if (t->index == 3) return 19; // for test - - blockIndex = p->blockIndex++; - - // PRF(printf("\ncanRead\n")) - - res = MtDec_Progress_GetError_Spec(p, 0, 0, blockIndex, &wasInterrupted); - - finish = p->readWasFinished; - needCode = False; - needWrite = False; - isAllocError = False; - overflow = False; - - inDataSize_Start = 0; - inDataSize = 0; - // inDataSize_Full = 0; - - if (res == SZ_OK && !wasInterrupted) - { - // if (p->inStream) - { - CMtDecBufLink *prev = NULL; - CMtDecBufLink *link = (CMtDecBufLink *)t->inBuf; - size_t crossSize = p->crossEnd - p->crossStart; - - PRF(printf("\ncrossSize = %d\n", crossSize)); - - for (;;) - { - if (!link) - { - link = (CMtDecBufLink *)ISzAlloc_Alloc(p->alloc, MTDEC__LINK_DATA_OFFSET + p->inBufSize); - if (!link) - { - finish = True; - // p->allocError_for_Read_BlockIndex = blockIndex; - isAllocError = True; - break; - } - link->next = NULL; - if (prev) - { - // static unsigned g_num = 0; - // printf("\n%6d : %x", ++g_num, (unsigned)(size_t)((Byte *)link - (Byte *)prev)); - prev->next = link; - } - else - t->inBuf = (void *)link; - } - - { - Byte *data = MTDEC__DATA_PTR_FROM_LINK(link); - Byte *parseData = data; - size_t size; - - if (crossSize != 0) - { - inDataSize = crossSize; - // inDataSize_Full = inDataSize; - inDataSize_Start = crossSize; - size = crossSize; - parseData = MTDEC__DATA_PTR_FROM_LINK(p->crossBlock) + p->crossStart; - PRF(printf("\ncross : crossStart = %7d crossEnd = %7d finish = %1d", - (int)p->crossStart, (int)p->crossEnd, (int)finish)); - } - else - { - size = p->inBufSize; - - res = FullRead(p->inStream, data, &size); - - // size = 10; // test - - inDataSize += size; - // inDataSize_Full = inDataSize; - if (!prev) - inDataSize_Start = size; - - p->readProcessed += size; - finish = (size != p->inBufSize); - if (finish) - p->readWasFinished = True; - - // res = E_INVALIDARG; // test - - if (res != SZ_OK) - { - // PRF(printf("\nRead error = %d\n", res)) - // we want to decode all data before error - p->readRes = res; - // p->readError_BlockIndex = blockIndex; - p->readWasFinished = True; - finish = True; - res = SZ_OK; - // break; - } - - if (inDataSize - inPrev >= MTDEC_ProgessStep) - { - res = MtDec_Progress_GetError_Spec(p, 0, 0, blockIndex, &wasInterrupted); - if (res != SZ_OK || wasInterrupted) - break; - inPrev = inDataSize; - } - } - - { - CMtDecCallbackInfo parse; - - parse.startCall = (prev == NULL); - parse.src = parseData; - parse.srcSize = size; - parse.srcFinished = finish; - parse.canCreateNewThread = True; - - PRF(printf("\nParse size = %d\n", (unsigned)size)); - - p->mtCallback->Parse(p->mtCallbackObject, t->index, &parse); - - PRF(printf(" Parse processed = %d, state = %d \n", (unsigned)parse.srcSize, (unsigned)parse.state)); - - needWrite = True; - canCreateNewThread = parse.canCreateNewThread; - - // printf("\n\n%12I64u %12I64u", (UInt64)p->mtProgress.totalInSize, (UInt64)p->mtProgress.totalOutSize); - - if ( - // parseRes != SZ_OK || - // inDataSize - (size - parse.srcSize) > p->inBlockMax - // || - parse.state == MTDEC_PARSE_OVERFLOW - // || wasInterrupted - ) - { - // Overflow or Parse error - switch from MT decoding to ST decoding - finish = True; - overflow = True; - - { - PRF(printf("\n Overflow")); - // PRF(printf("\nisBlockFinished = %d", (unsigned)parse.blockWasFinished)); - PRF(printf("\n inDataSize = %d", (unsigned)inDataSize)); - } - - if (crossSize != 0) - memcpy(data, parseData, size); - p->crossStart = 0; - p->crossEnd = 0; - break; - } - - if (crossSize != 0) - { - memcpy(data, parseData, parse.srcSize); - p->crossStart += parse.srcSize; - } - - if (parse.state != MTDEC_PARSE_CONTINUE || finish) - { - // we don't need to parse in current thread anymore - - if (parse.state == MTDEC_PARSE_END) - finish = True; - - needCode = True; - // p->crossFinished = finish; - - if (parse.srcSize == size) - { - // full parsed - no cross transfer - p->crossStart = 0; - p->crossEnd = 0; - break; - } - - if (parse.state == MTDEC_PARSE_END) - { - afterEndData = parseData + parse.srcSize; - afterEndData_Size = size - parse.srcSize; - if (crossSize != 0) - afterEndData_IsCross = True; - // we reduce data size to required bytes (parsed only) - inDataSize -= afterEndData_Size; - if (!prev) - inDataSize_Start = parse.srcSize; - break; - } - - { - // partial parsed - need cross transfer - if (crossSize != 0) - inDataSize = parse.srcSize; // it's only parsed now - else - { - // partial parsed - is not in initial cross block - we need to copy new data to cross block - Byte *cr = MtDec_GetCrossBuff(p); - if (!cr) - { - { - PRF(printf("\ncross alloc error error\n")); - // res = SZ_ERROR_MEM; - finish = True; - // p->allocError_for_Read_BlockIndex = blockIndex; - isAllocError = True; - break; - } - } - - { - size_t crSize = size - parse.srcSize; - inDataSize -= crSize; - p->crossEnd = crSize; - p->crossStart = 0; - memcpy(cr, parseData + parse.srcSize, crSize); - } - } - - // inDataSize_Full = inDataSize; - if (!prev) - inDataSize_Start = parse.srcSize; // it's partial size (parsed only) - - finish = False; - break; - } - } - - if (parse.srcSize != size) - { - res = SZ_ERROR_FAIL; - PRF(printf("\nfinished error SZ_ERROR_FAIL = %d\n", res)); - break; - } - } - } - - prev = link; - link = link->next; - - if (crossSize != 0) - { - crossSize = 0; - p->crossStart = 0; - p->crossEnd = 0; - } - } - } - - if (res == SZ_OK) - res = MtDec_GetError_Spec(p, blockIndex, &wasInterrupted); - } - - codeRes = SZ_OK; - - if (res == SZ_OK && needCode && !wasInterrupted) - { - codeRes = p->mtCallback->PreCode(p->mtCallbackObject, t->index); - if (codeRes != SZ_OK) - { - needCode = False; - finish = True; - // SZ_ERROR_MEM is expected error here. - // if (codeRes == SZ_ERROR_MEM) - we will try single-thread decoding later. - // if (codeRes != SZ_ERROR_MEM) - we can stop decoding or try single-thread decoding. - } - } - - if (res != SZ_OK || wasInterrupted) - finish = True; - - nextThread = NULL; - threadingErrorSRes = SZ_OK; - - if (!finish) - { - if (p->numStartedThreads < p->numStartedThreads_Limit && canCreateNewThread) - { - SRes res2 = MtDecThread_CreateAndStart(&p->threads[p->numStartedThreads]); - if (res2 == SZ_OK) - { - // if (p->numStartedThreads % 1000 == 0) PRF(printf("\n numStartedThreads=%d\n", p->numStartedThreads)); - p->numStartedThreads++; - } - else - { - PRF(printf("\nERROR: numStartedThreads=%d\n", p->numStartedThreads)); - if (p->numStartedThreads == 1) - { - // if only one thread is possible, we leave muti-threading code - finish = True; - needCode = False; - threadingErrorSRes = res2; - } - else - p->numStartedThreads_Limit = p->numStartedThreads; - } - } - - if (!finish) - { - unsigned nextIndex = t->index + 1; - nextThread = &p->threads[nextIndex >= p->numStartedThreads ? 0 : nextIndex]; - RINOK_THREAD(Event_Set(&nextThread->canRead)) - // We have started executing for new iteration (with next thread) - // And that next thread now is responsible for possible exit from decoding (threading_code) - } - } - - // each call of Event_Set(&nextThread->canRead) must be followed by call of Event_Set(&nextThread->canWrite) - // if ( !finish ) we must call Event_Set(&nextThread->canWrite) in any case - // if ( finish ) we switch to single-thread mode and there are 2 ways at the end of current iteration (current block): - // - if (needContinue) after Write(&needContinue), we restore decoding with new iteration - // - otherwise we stop decoding and exit from ThreadFunc2() - - // Don't change (finish) variable in the further code - - - // ---------- CODE ---------- - - inPrev = 0; - outPrev = 0; - inCodePos = 0; - outCodePos = 0; - - if (res == SZ_OK && needCode && codeRes == SZ_OK) - { - BoolInt isStartBlock = True; - CMtDecBufLink *link = (CMtDecBufLink *)t->inBuf; - - for (;;) - { - size_t inSize; - int stop; - - if (isStartBlock) - inSize = inDataSize_Start; - else - { - UInt64 rem = inDataSize - inCodePos; - inSize = p->inBufSize; - if (inSize > rem) - inSize = (size_t)rem; - } - - inCodePos += inSize; - stop = True; - - codeRes = p->mtCallback->Code(p->mtCallbackObject, t->index, - (const Byte *)MTDEC__DATA_PTR_FROM_LINK(link), inSize, - (inCodePos == inDataSize), // srcFinished - &inCodePos, &outCodePos, &stop); - - if (codeRes != SZ_OK) - { - PRF(printf("\nCode Interrupt error = %x\n", codeRes)); - // we interrupt only later blocks - MtDec_Interrupt(p, blockIndex); - break; - } - - if (stop || inCodePos == inDataSize) - break; - - { - const UInt64 inDelta = inCodePos - inPrev; - const UInt64 outDelta = outCodePos - outPrev; - if (inDelta >= MTDEC_ProgessStep || outDelta >= MTDEC_ProgessStep) - { - // Sleep(1); - res = MtDec_Progress_GetError_Spec(p, inDelta, outDelta, blockIndex, &wasInterrupted); - if (res != SZ_OK || wasInterrupted) - break; - inPrev = inCodePos; - outPrev = outCodePos; - } - } - - link = link->next; - isStartBlock = False; - } - } - - - // ---------- WRITE ---------- - - RINOK_THREAD(Event_Wait(&t->canWrite)); - - { - BoolInt isErrorMode = False; - BoolInt canRecode = True; - BoolInt needWriteToStream = needWrite; - - if (p->exitThread) return 0; // it's never executed in normal cases - - if (p->wasInterrupted) - wasInterrupted = True; - else - { - if (codeRes != SZ_OK) // || !needCode // check it !!! - { - p->wasInterrupted = True; - p->codeRes = codeRes; - if (codeRes == SZ_ERROR_MEM) - isAllocError = True; - } - - if (threadingErrorSRes) - { - p->wasInterrupted = True; - p->threadingErrorSRes = threadingErrorSRes; - needWriteToStream = False; - } - if (isAllocError) - { - p->wasInterrupted = True; - p->isAllocError = True; - needWriteToStream = False; - } - if (overflow) - { - p->wasInterrupted = True; - p->overflow = True; - needWriteToStream = False; - } - } - - if (needCode) - { - if (wasInterrupted) - { - inCodePos = 0; - outCodePos = 0; - } - { - const UInt64 inDelta = inCodePos - inPrev; - const UInt64 outDelta = outCodePos - outPrev; - // if (inDelta != 0 || outDelta != 0) - res = MtProgress_ProgressAdd(&p->mtProgress, inDelta, outDelta); - } - } - - needContinue = (!finish); - - // if (res == SZ_OK && needWrite && !wasInterrupted) - if (needWrite) - { - // p->inProcessed += inCodePos; - - PRF(printf("\n--Write afterSize = %d\n", (unsigned)afterEndData_Size)); - - res = p->mtCallback->Write(p->mtCallbackObject, t->index, - res == SZ_OK && needWriteToStream && !wasInterrupted, // needWrite - afterEndData, afterEndData_Size, afterEndData_IsCross, - &needContinue, - &canRecode); - - // res = SZ_ERROR_FAIL; // for test - - PRF(printf("\nAfter Write needContinue = %d\n", (unsigned)needContinue)); - PRF(printf("\nprocessed = %d\n", (unsigned)p->inProcessed)); - - if (res != SZ_OK) - { - PRF(printf("\nWrite error = %d\n", res)); - isErrorMode = True; - p->wasInterrupted = True; - } - if (res != SZ_OK - || (!needContinue && !finish)) - { - PRF(printf("\nWrite Interrupt error = %x\n", res)); - MtDec_Interrupt(p, blockIndex); - } - } - - if (canRecode) - if (!needCode - || res != SZ_OK - || p->wasInterrupted - || codeRes != SZ_OK - || wasInterrupted - || p->numFilledThreads != 0 - || isErrorMode) - { - if (p->numFilledThreads == 0) - p->filledThreadStart = t->index; - if (inDataSize != 0 || !finish) - { - t->inDataSize_Start = inDataSize_Start; - t->inDataSize = inDataSize; - p->numFilledThreads++; - } - PRF(printf("\np->numFilledThreads = %d\n", p->numFilledThreads)); - PRF(printf("p->filledThreadStart = %d\n", p->filledThreadStart)); - } - - if (!finish) - { - RINOK_THREAD(Event_Set(&nextThread->canWrite)); - } - else - { - if (needContinue) - { - // we restore decoding with new iteration - RINOK_THREAD(Event_Set(&p->threads[0].canWrite)); - } - else - { - // we exit from decoding - if (t->index == 0) - return SZ_OK; - p->exitThread = True; - } - RINOK_THREAD(Event_Set(&p->threads[0].canRead)); - } - } - } -} - -#ifdef _WIN32 -#define USE_ALLOCA -#endif - -#ifdef USE_ALLOCA -#ifdef _WIN32 -#include -#else -#include -#endif -#endif - - -static THREAD_FUNC_DECL ThreadFunc1(void *pp) -{ - WRes res; - - CMtDecThread *t = (CMtDecThread *)pp; - CMtDec *p; - - // fprintf(stdout, "\n%d = %p\n", t->index, &t); - - res = ThreadFunc2(t); - p = t->mtDec; - if (res == 0) - return (THREAD_FUNC_RET_TYPE)(UINT_PTR)p->exitThreadWRes; - { - // it's unexpected situation for some threading function error - if (p->exitThreadWRes == 0) - p->exitThreadWRes = res; - PRF(printf("\nthread exit error = %d\n", res)); - p->exitThread = True; - Event_Set(&p->threads[0].canRead); - Event_Set(&p->threads[0].canWrite); - MtProgress_SetError(&p->mtProgress, MY_SRes_HRESULT_FROM_WRes(res)); - } - return (THREAD_FUNC_RET_TYPE)(UINT_PTR)res; -} - -static MY_NO_INLINE THREAD_FUNC_DECL ThreadFunc(void *pp) -{ - #ifdef USE_ALLOCA - CMtDecThread *t = (CMtDecThread *)pp; - // fprintf(stderr, "\n%d = %p - before", t->index, &t); - t->allocaPtr = alloca(t->index * 128); - #endif - return ThreadFunc1(pp); -} - - -int MtDec_PrepareRead(CMtDec *p) -{ - if (p->crossBlock && p->crossStart == p->crossEnd) - { - ISzAlloc_Free(p->alloc, p->crossBlock); - p->crossBlock = NULL; - } - - { - unsigned i; - for (i = 0; i < MTDEC__THREADS_MAX; i++) - if (i > p->numStartedThreads - || p->numFilledThreads <= - (i >= p->filledThreadStart ? - i - p->filledThreadStart : - i + p->numStartedThreads - p->filledThreadStart)) - MtDecThread_FreeInBufs(&p->threads[i]); - } - - return (p->numFilledThreads != 0) || (p->crossStart != p->crossEnd); -} - - -const Byte *MtDec_Read(CMtDec *p, size_t *inLim) -{ - while (p->numFilledThreads != 0) - { - CMtDecThread *t = &p->threads[p->filledThreadStart]; - - if (*inLim != 0) - { - { - void *link = t->inBuf; - void *next = ((CMtDecBufLink *)link)->next; - ISzAlloc_Free(p->alloc, link); - t->inBuf = next; - } - - if (t->inDataSize == 0) - { - MtDecThread_FreeInBufs(t); - if (--p->numFilledThreads == 0) - break; - if (++p->filledThreadStart == p->numStartedThreads) - p->filledThreadStart = 0; - t = &p->threads[p->filledThreadStart]; - } - } - - { - size_t lim = t->inDataSize_Start; - if (lim != 0) - t->inDataSize_Start = 0; - else - { - UInt64 rem = t->inDataSize; - lim = p->inBufSize; - if (lim > rem) - lim = (size_t)rem; - } - t->inDataSize -= lim; - *inLim = lim; - return (const Byte *)MTDEC__DATA_PTR_FROM_LINK(t->inBuf); - } - } - - { - size_t crossSize = p->crossEnd - p->crossStart; - if (crossSize != 0) - { - const Byte *data = MTDEC__DATA_PTR_FROM_LINK(p->crossBlock) + p->crossStart; - *inLim = crossSize; - p->crossStart = 0; - p->crossEnd = 0; - return data; - } - *inLim = 0; - if (p->crossBlock) - { - ISzAlloc_Free(p->alloc, p->crossBlock); - p->crossBlock = NULL; - } - return NULL; - } -} - - -void MtDec_Construct(CMtDec *p) -{ - unsigned i; - - p->inBufSize = (size_t)1 << 18; - - p->numThreadsMax = 0; - - p->inStream = NULL; - - // p->inData = NULL; - // p->inDataSize = 0; - - p->crossBlock = NULL; - p->crossStart = 0; - p->crossEnd = 0; - - p->numFilledThreads = 0; - - p->progress = NULL; - p->alloc = NULL; - - p->mtCallback = NULL; - p->mtCallbackObject = NULL; - - p->allocatedBufsSize = 0; - - for (i = 0; i < MTDEC__THREADS_MAX; i++) - { - CMtDecThread *t = &p->threads[i]; - t->mtDec = p; - t->index = i; - t->inBuf = NULL; - Event_Construct(&t->canRead); - Event_Construct(&t->canWrite); - Thread_Construct(&t->thread); - } - - // Event_Construct(&p->finishedEvent); - - CriticalSection_Init(&p->mtProgress.cs); -} - - -static void MtDec_Free(CMtDec *p) -{ - unsigned i; - - p->exitThread = True; - - for (i = 0; i < MTDEC__THREADS_MAX; i++) - MtDecThread_Destruct(&p->threads[i]); - - // Event_Close(&p->finishedEvent); - - if (p->crossBlock) - { - ISzAlloc_Free(p->alloc, p->crossBlock); - p->crossBlock = NULL; - } -} - - -void MtDec_Destruct(CMtDec *p) -{ - MtDec_Free(p); - - CriticalSection_Delete(&p->mtProgress.cs); -} - - -SRes MtDec_Code(CMtDec *p) -{ - unsigned i; - - p->inProcessed = 0; - - p->blockIndex = 1; // it must be larger than not_defined index (0) - p->isAllocError = False; - p->overflow = False; - p->threadingErrorSRes = SZ_OK; - - p->needContinue = True; - - p->readWasFinished = False; - p->needInterrupt = False; - p->interruptIndex = (UInt64)(Int64)-1; - - p->readProcessed = 0; - p->readRes = SZ_OK; - p->codeRes = SZ_OK; - p->wasInterrupted = False; - - p->crossStart = 0; - p->crossEnd = 0; - - p->filledThreadStart = 0; - p->numFilledThreads = 0; - - { - unsigned numThreads = p->numThreadsMax; - if (numThreads > MTDEC__THREADS_MAX) - numThreads = MTDEC__THREADS_MAX; - p->numStartedThreads_Limit = numThreads; - p->numStartedThreads = 0; - } - - if (p->inBufSize != p->allocatedBufsSize) - { - for (i = 0; i < MTDEC__THREADS_MAX; i++) - { - CMtDecThread *t = &p->threads[i]; - if (t->inBuf) - MtDecThread_FreeInBufs(t); - } - if (p->crossBlock) - { - ISzAlloc_Free(p->alloc, p->crossBlock); - p->crossBlock = NULL; - } - - p->allocatedBufsSize = p->inBufSize; - } - - MtProgress_Init(&p->mtProgress, p->progress); - - // RINOK_THREAD(ArEvent_OptCreate_And_Reset(&p->finishedEvent)); - p->exitThread = False; - p->exitThreadWRes = 0; - - { - WRes wres; - SRes sres; - CMtDecThread *nextThread = &p->threads[p->numStartedThreads++]; - // wres = MtDecThread_CreateAndStart(nextThread); - wres = MtDecThread_CreateEvents(nextThread); - if (wres == 0) { wres = Event_Set(&nextThread->canWrite); - if (wres == 0) { wres = Event_Set(&nextThread->canRead); - if (wres == 0) { THREAD_FUNC_RET_TYPE res = ThreadFunc(nextThread); - wres = (WRes)(UINT_PTR)res; - if (wres != 0) - { - p->needContinue = False; - MtDec_CloseThreads(p); - }}}} - - // wres = 17; // for test - // wres = Event_Wait(&p->finishedEvent); - - sres = MY_SRes_HRESULT_FROM_WRes(wres); - - if (sres != 0) - p->threadingErrorSRes = sres; - - if ( - // wres == 0 - // wres != 0 - // || p->mtc.codeRes == SZ_ERROR_MEM - p->isAllocError - || p->threadingErrorSRes != SZ_OK - || p->overflow) - { - // p->needContinue = True; - } - else - p->needContinue = False; - - if (p->needContinue) - return SZ_OK; - - // if (sres != SZ_OK) - return sres; - // return SZ_ERROR_FAIL; - } -} - -#endif +/* MtDec.c -- Multi-thread Decoder +2021-12-21 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +// #define SHOW_DEBUG_INFO + +// #include +#include + +#ifdef SHOW_DEBUG_INFO +#include +#endif + +#include "MtDec.h" + +#ifndef _7ZIP_ST + +#ifdef SHOW_DEBUG_INFO +#define PRF(x) x +#else +#define PRF(x) +#endif + +#define PRF_STR_INT(s, d) PRF(printf("\n" s " %d\n", (unsigned)d)) + +void MtProgress_Init(CMtProgress *p, ICompressProgress *progress) +{ + p->progress = progress; + p->res = SZ_OK; + p->totalInSize = 0; + p->totalOutSize = 0; +} + + +SRes MtProgress_Progress_ST(CMtProgress *p) +{ + if (p->res == SZ_OK && p->progress) + if (ICompressProgress_Progress(p->progress, p->totalInSize, p->totalOutSize) != SZ_OK) + p->res = SZ_ERROR_PROGRESS; + return p->res; +} + + +SRes MtProgress_ProgressAdd(CMtProgress *p, UInt64 inSize, UInt64 outSize) +{ + SRes res; + CriticalSection_Enter(&p->cs); + + p->totalInSize += inSize; + p->totalOutSize += outSize; + if (p->res == SZ_OK && p->progress) + if (ICompressProgress_Progress(p->progress, p->totalInSize, p->totalOutSize) != SZ_OK) + p->res = SZ_ERROR_PROGRESS; + res = p->res; + + CriticalSection_Leave(&p->cs); + return res; +} + + +SRes MtProgress_GetError(CMtProgress *p) +{ + SRes res; + CriticalSection_Enter(&p->cs); + res = p->res; + CriticalSection_Leave(&p->cs); + return res; +} + + +void MtProgress_SetError(CMtProgress *p, SRes res) +{ + CriticalSection_Enter(&p->cs); + if (p->res == SZ_OK) + p->res = res; + CriticalSection_Leave(&p->cs); +} + + +#define RINOK_THREAD(x) RINOK_WRes(x) + + +static WRes ArEvent_OptCreate_And_Reset(CEvent *p) +{ + if (Event_IsCreated(p)) + return Event_Reset(p); + return AutoResetEvent_CreateNotSignaled(p); +} + + +struct __CMtDecBufLink +{ + struct __CMtDecBufLink *next; + void *pad[3]; +}; + +typedef struct __CMtDecBufLink CMtDecBufLink; + +#define MTDEC__LINK_DATA_OFFSET sizeof(CMtDecBufLink) +#define MTDEC__DATA_PTR_FROM_LINK(link) ((Byte *)(link) + MTDEC__LINK_DATA_OFFSET) + + + +static THREAD_FUNC_DECL ThreadFunc(void *pp); + + +static WRes MtDecThread_CreateEvents(CMtDecThread *t) +{ + WRes wres = ArEvent_OptCreate_And_Reset(&t->canWrite); + if (wres == 0) + { + wres = ArEvent_OptCreate_And_Reset(&t->canRead); + if (wres == 0) + return SZ_OK; + } + return wres; +} + + +static SRes MtDecThread_CreateAndStart(CMtDecThread *t) +{ + WRes wres = MtDecThread_CreateEvents(t); + // wres = 17; // for test + if (wres == 0) + { + if (Thread_WasCreated(&t->thread)) + return SZ_OK; + wres = Thread_Create(&t->thread, ThreadFunc, t); + if (wres == 0) + return SZ_OK; + } + return MY_SRes_HRESULT_FROM_WRes(wres); +} + + +void MtDecThread_FreeInBufs(CMtDecThread *t) +{ + if (t->inBuf) + { + void *link = t->inBuf; + t->inBuf = NULL; + do + { + void *next = ((CMtDecBufLink *)link)->next; + ISzAlloc_Free(t->mtDec->alloc, link); + link = next; + } + while (link); + } +} + + +static void MtDecThread_CloseThread(CMtDecThread *t) +{ + if (Thread_WasCreated(&t->thread)) + { + Event_Set(&t->canWrite); /* we can disable it. There are no threads waiting canWrite in normal cases */ + Event_Set(&t->canRead); + Thread_Wait_Close(&t->thread); + } + + Event_Close(&t->canRead); + Event_Close(&t->canWrite); +} + +static void MtDec_CloseThreads(CMtDec *p) +{ + unsigned i; + for (i = 0; i < MTDEC__THREADS_MAX; i++) + MtDecThread_CloseThread(&p->threads[i]); +} + +static void MtDecThread_Destruct(CMtDecThread *t) +{ + MtDecThread_CloseThread(t); + MtDecThread_FreeInBufs(t); +} + + + +static SRes FullRead(ISeqInStream *stream, Byte *data, size_t *processedSize) +{ + size_t size = *processedSize; + *processedSize = 0; + while (size != 0) + { + size_t cur = size; + SRes res = ISeqInStream_Read(stream, data, &cur); + *processedSize += cur; + data += cur; + size -= cur; + RINOK(res); + if (cur == 0) + return SZ_OK; + } + return SZ_OK; +} + + +static SRes MtDec_GetError_Spec(CMtDec *p, UInt64 interruptIndex, BoolInt *wasInterrupted) +{ + SRes res; + CriticalSection_Enter(&p->mtProgress.cs); + *wasInterrupted = (p->needInterrupt && interruptIndex > p->interruptIndex); + res = p->mtProgress.res; + CriticalSection_Leave(&p->mtProgress.cs); + return res; +} + +static SRes MtDec_Progress_GetError_Spec(CMtDec *p, UInt64 inSize, UInt64 outSize, UInt64 interruptIndex, BoolInt *wasInterrupted) +{ + SRes res; + CriticalSection_Enter(&p->mtProgress.cs); + + p->mtProgress.totalInSize += inSize; + p->mtProgress.totalOutSize += outSize; + if (p->mtProgress.res == SZ_OK && p->mtProgress.progress) + if (ICompressProgress_Progress(p->mtProgress.progress, p->mtProgress.totalInSize, p->mtProgress.totalOutSize) != SZ_OK) + p->mtProgress.res = SZ_ERROR_PROGRESS; + + *wasInterrupted = (p->needInterrupt && interruptIndex > p->interruptIndex); + res = p->mtProgress.res; + + CriticalSection_Leave(&p->mtProgress.cs); + + return res; +} + +static void MtDec_Interrupt(CMtDec *p, UInt64 interruptIndex) +{ + CriticalSection_Enter(&p->mtProgress.cs); + if (!p->needInterrupt || interruptIndex < p->interruptIndex) + { + p->interruptIndex = interruptIndex; + p->needInterrupt = True; + } + CriticalSection_Leave(&p->mtProgress.cs); +} + +Byte *MtDec_GetCrossBuff(CMtDec *p) +{ + Byte *cr = p->crossBlock; + if (!cr) + { + cr = (Byte *)ISzAlloc_Alloc(p->alloc, MTDEC__LINK_DATA_OFFSET + p->inBufSize); + if (!cr) + return NULL; + p->crossBlock = cr; + } + return MTDEC__DATA_PTR_FROM_LINK(cr); +} + + +/* + ThreadFunc2() returns: + 0 - in all normal cases (even for stream error or memory allocation error) + (!= 0) - WRes error return by system threading function +*/ + +// #define MTDEC_ProgessStep (1 << 22) +#define MTDEC_ProgessStep (1 << 0) + +static WRes ThreadFunc2(CMtDecThread *t) +{ + CMtDec *p = t->mtDec; + + PRF_STR_INT("ThreadFunc2", t->index); + + // SetThreadAffinityMask(GetCurrentThread(), 1 << t->index); + + for (;;) + { + SRes res, codeRes; + BoolInt wasInterrupted, isAllocError, overflow, finish; + SRes threadingErrorSRes; + BoolInt needCode, needWrite, needContinue; + + size_t inDataSize_Start; + UInt64 inDataSize; + // UInt64 inDataSize_Full; + + UInt64 blockIndex; + + UInt64 inPrev = 0; + UInt64 outPrev = 0; + UInt64 inCodePos; + UInt64 outCodePos; + + Byte *afterEndData = NULL; + size_t afterEndData_Size = 0; + BoolInt afterEndData_IsCross = False; + + BoolInt canCreateNewThread = False; + // CMtDecCallbackInfo parse; + CMtDecThread *nextThread; + + PRF_STR_INT("=============== Event_Wait(&t->canRead)", t->index); + + RINOK_THREAD(Event_Wait(&t->canRead)); + if (p->exitThread) + return 0; + + PRF_STR_INT("after Event_Wait(&t->canRead)", t->index); + + // if (t->index == 3) return 19; // for test + + blockIndex = p->blockIndex++; + + // PRF(printf("\ncanRead\n")) + + res = MtDec_Progress_GetError_Spec(p, 0, 0, blockIndex, &wasInterrupted); + + finish = p->readWasFinished; + needCode = False; + needWrite = False; + isAllocError = False; + overflow = False; + + inDataSize_Start = 0; + inDataSize = 0; + // inDataSize_Full = 0; + + if (res == SZ_OK && !wasInterrupted) + { + // if (p->inStream) + { + CMtDecBufLink *prev = NULL; + CMtDecBufLink *link = (CMtDecBufLink *)t->inBuf; + size_t crossSize = p->crossEnd - p->crossStart; + + PRF(printf("\ncrossSize = %d\n", crossSize)); + + for (;;) + { + if (!link) + { + link = (CMtDecBufLink *)ISzAlloc_Alloc(p->alloc, MTDEC__LINK_DATA_OFFSET + p->inBufSize); + if (!link) + { + finish = True; + // p->allocError_for_Read_BlockIndex = blockIndex; + isAllocError = True; + break; + } + link->next = NULL; + if (prev) + { + // static unsigned g_num = 0; + // printf("\n%6d : %x", ++g_num, (unsigned)(size_t)((Byte *)link - (Byte *)prev)); + prev->next = link; + } + else + t->inBuf = (void *)link; + } + + { + Byte *data = MTDEC__DATA_PTR_FROM_LINK(link); + Byte *parseData = data; + size_t size; + + if (crossSize != 0) + { + inDataSize = crossSize; + // inDataSize_Full = inDataSize; + inDataSize_Start = crossSize; + size = crossSize; + parseData = MTDEC__DATA_PTR_FROM_LINK(p->crossBlock) + p->crossStart; + PRF(printf("\ncross : crossStart = %7d crossEnd = %7d finish = %1d", + (int)p->crossStart, (int)p->crossEnd, (int)finish)); + } + else + { + size = p->inBufSize; + + res = FullRead(p->inStream, data, &size); + + // size = 10; // test + + inDataSize += size; + // inDataSize_Full = inDataSize; + if (!prev) + inDataSize_Start = size; + + p->readProcessed += size; + finish = (size != p->inBufSize); + if (finish) + p->readWasFinished = True; + + // res = E_INVALIDARG; // test + + if (res != SZ_OK) + { + // PRF(printf("\nRead error = %d\n", res)) + // we want to decode all data before error + p->readRes = res; + // p->readError_BlockIndex = blockIndex; + p->readWasFinished = True; + finish = True; + res = SZ_OK; + // break; + } + + if (inDataSize - inPrev >= MTDEC_ProgessStep) + { + res = MtDec_Progress_GetError_Spec(p, 0, 0, blockIndex, &wasInterrupted); + if (res != SZ_OK || wasInterrupted) + break; + inPrev = inDataSize; + } + } + + { + CMtDecCallbackInfo parse; + + parse.startCall = (prev == NULL); + parse.src = parseData; + parse.srcSize = size; + parse.srcFinished = finish; + parse.canCreateNewThread = True; + + PRF(printf("\nParse size = %d\n", (unsigned)size)); + + p->mtCallback->Parse(p->mtCallbackObject, t->index, &parse); + + PRF(printf(" Parse processed = %d, state = %d \n", (unsigned)parse.srcSize, (unsigned)parse.state)); + + needWrite = True; + canCreateNewThread = parse.canCreateNewThread; + + // printf("\n\n%12I64u %12I64u", (UInt64)p->mtProgress.totalInSize, (UInt64)p->mtProgress.totalOutSize); + + if ( + // parseRes != SZ_OK || + // inDataSize - (size - parse.srcSize) > p->inBlockMax + // || + parse.state == MTDEC_PARSE_OVERFLOW + // || wasInterrupted + ) + { + // Overflow or Parse error - switch from MT decoding to ST decoding + finish = True; + overflow = True; + + { + PRF(printf("\n Overflow")); + // PRF(printf("\nisBlockFinished = %d", (unsigned)parse.blockWasFinished)); + PRF(printf("\n inDataSize = %d", (unsigned)inDataSize)); + } + + if (crossSize != 0) + memcpy(data, parseData, size); + p->crossStart = 0; + p->crossEnd = 0; + break; + } + + if (crossSize != 0) + { + memcpy(data, parseData, parse.srcSize); + p->crossStart += parse.srcSize; + } + + if (parse.state != MTDEC_PARSE_CONTINUE || finish) + { + // we don't need to parse in current thread anymore + + if (parse.state == MTDEC_PARSE_END) + finish = True; + + needCode = True; + // p->crossFinished = finish; + + if (parse.srcSize == size) + { + // full parsed - no cross transfer + p->crossStart = 0; + p->crossEnd = 0; + break; + } + + if (parse.state == MTDEC_PARSE_END) + { + afterEndData = parseData + parse.srcSize; + afterEndData_Size = size - parse.srcSize; + if (crossSize != 0) + afterEndData_IsCross = True; + // we reduce data size to required bytes (parsed only) + inDataSize -= afterEndData_Size; + if (!prev) + inDataSize_Start = parse.srcSize; + break; + } + + { + // partial parsed - need cross transfer + if (crossSize != 0) + inDataSize = parse.srcSize; // it's only parsed now + else + { + // partial parsed - is not in initial cross block - we need to copy new data to cross block + Byte *cr = MtDec_GetCrossBuff(p); + if (!cr) + { + { + PRF(printf("\ncross alloc error error\n")); + // res = SZ_ERROR_MEM; + finish = True; + // p->allocError_for_Read_BlockIndex = blockIndex; + isAllocError = True; + break; + } + } + + { + size_t crSize = size - parse.srcSize; + inDataSize -= crSize; + p->crossEnd = crSize; + p->crossStart = 0; + memcpy(cr, parseData + parse.srcSize, crSize); + } + } + + // inDataSize_Full = inDataSize; + if (!prev) + inDataSize_Start = parse.srcSize; // it's partial size (parsed only) + + finish = False; + break; + } + } + + if (parse.srcSize != size) + { + res = SZ_ERROR_FAIL; + PRF(printf("\nfinished error SZ_ERROR_FAIL = %d\n", res)); + break; + } + } + } + + prev = link; + link = link->next; + + if (crossSize != 0) + { + crossSize = 0; + p->crossStart = 0; + p->crossEnd = 0; + } + } + } + + if (res == SZ_OK) + res = MtDec_GetError_Spec(p, blockIndex, &wasInterrupted); + } + + codeRes = SZ_OK; + + if (res == SZ_OK && needCode && !wasInterrupted) + { + codeRes = p->mtCallback->PreCode(p->mtCallbackObject, t->index); + if (codeRes != SZ_OK) + { + needCode = False; + finish = True; + // SZ_ERROR_MEM is expected error here. + // if (codeRes == SZ_ERROR_MEM) - we will try single-thread decoding later. + // if (codeRes != SZ_ERROR_MEM) - we can stop decoding or try single-thread decoding. + } + } + + if (res != SZ_OK || wasInterrupted) + finish = True; + + nextThread = NULL; + threadingErrorSRes = SZ_OK; + + if (!finish) + { + if (p->numStartedThreads < p->numStartedThreads_Limit && canCreateNewThread) + { + SRes res2 = MtDecThread_CreateAndStart(&p->threads[p->numStartedThreads]); + if (res2 == SZ_OK) + { + // if (p->numStartedThreads % 1000 == 0) PRF(printf("\n numStartedThreads=%d\n", p->numStartedThreads)); + p->numStartedThreads++; + } + else + { + PRF(printf("\nERROR: numStartedThreads=%d\n", p->numStartedThreads)); + if (p->numStartedThreads == 1) + { + // if only one thread is possible, we leave muti-threading code + finish = True; + needCode = False; + threadingErrorSRes = res2; + } + else + p->numStartedThreads_Limit = p->numStartedThreads; + } + } + + if (!finish) + { + unsigned nextIndex = t->index + 1; + nextThread = &p->threads[nextIndex >= p->numStartedThreads ? 0 : nextIndex]; + RINOK_THREAD(Event_Set(&nextThread->canRead)) + // We have started executing for new iteration (with next thread) + // And that next thread now is responsible for possible exit from decoding (threading_code) + } + } + + // each call of Event_Set(&nextThread->canRead) must be followed by call of Event_Set(&nextThread->canWrite) + // if ( !finish ) we must call Event_Set(&nextThread->canWrite) in any case + // if ( finish ) we switch to single-thread mode and there are 2 ways at the end of current iteration (current block): + // - if (needContinue) after Write(&needContinue), we restore decoding with new iteration + // - otherwise we stop decoding and exit from ThreadFunc2() + + // Don't change (finish) variable in the further code + + + // ---------- CODE ---------- + + inPrev = 0; + outPrev = 0; + inCodePos = 0; + outCodePos = 0; + + if (res == SZ_OK && needCode && codeRes == SZ_OK) + { + BoolInt isStartBlock = True; + CMtDecBufLink *link = (CMtDecBufLink *)t->inBuf; + + for (;;) + { + size_t inSize; + int stop; + + if (isStartBlock) + inSize = inDataSize_Start; + else + { + UInt64 rem = inDataSize - inCodePos; + inSize = p->inBufSize; + if (inSize > rem) + inSize = (size_t)rem; + } + + inCodePos += inSize; + stop = True; + + codeRes = p->mtCallback->Code(p->mtCallbackObject, t->index, + (const Byte *)MTDEC__DATA_PTR_FROM_LINK(link), inSize, + (inCodePos == inDataSize), // srcFinished + &inCodePos, &outCodePos, &stop); + + if (codeRes != SZ_OK) + { + PRF(printf("\nCode Interrupt error = %x\n", codeRes)); + // we interrupt only later blocks + MtDec_Interrupt(p, blockIndex); + break; + } + + if (stop || inCodePos == inDataSize) + break; + + { + const UInt64 inDelta = inCodePos - inPrev; + const UInt64 outDelta = outCodePos - outPrev; + if (inDelta >= MTDEC_ProgessStep || outDelta >= MTDEC_ProgessStep) + { + // Sleep(1); + res = MtDec_Progress_GetError_Spec(p, inDelta, outDelta, blockIndex, &wasInterrupted); + if (res != SZ_OK || wasInterrupted) + break; + inPrev = inCodePos; + outPrev = outCodePos; + } + } + + link = link->next; + isStartBlock = False; + } + } + + + // ---------- WRITE ---------- + + RINOK_THREAD(Event_Wait(&t->canWrite)); + + { + BoolInt isErrorMode = False; + BoolInt canRecode = True; + BoolInt needWriteToStream = needWrite; + + if (p->exitThread) return 0; // it's never executed in normal cases + + if (p->wasInterrupted) + wasInterrupted = True; + else + { + if (codeRes != SZ_OK) // || !needCode // check it !!! + { + p->wasInterrupted = True; + p->codeRes = codeRes; + if (codeRes == SZ_ERROR_MEM) + isAllocError = True; + } + + if (threadingErrorSRes) + { + p->wasInterrupted = True; + p->threadingErrorSRes = threadingErrorSRes; + needWriteToStream = False; + } + if (isAllocError) + { + p->wasInterrupted = True; + p->isAllocError = True; + needWriteToStream = False; + } + if (overflow) + { + p->wasInterrupted = True; + p->overflow = True; + needWriteToStream = False; + } + } + + if (needCode) + { + if (wasInterrupted) + { + inCodePos = 0; + outCodePos = 0; + } + { + const UInt64 inDelta = inCodePos - inPrev; + const UInt64 outDelta = outCodePos - outPrev; + // if (inDelta != 0 || outDelta != 0) + res = MtProgress_ProgressAdd(&p->mtProgress, inDelta, outDelta); + } + } + + needContinue = (!finish); + + // if (res == SZ_OK && needWrite && !wasInterrupted) + if (needWrite) + { + // p->inProcessed += inCodePos; + + PRF(printf("\n--Write afterSize = %d\n", (unsigned)afterEndData_Size)); + + res = p->mtCallback->Write(p->mtCallbackObject, t->index, + res == SZ_OK && needWriteToStream && !wasInterrupted, // needWrite + afterEndData, afterEndData_Size, afterEndData_IsCross, + &needContinue, + &canRecode); + + // res = SZ_ERROR_FAIL; // for test + + PRF(printf("\nAfter Write needContinue = %d\n", (unsigned)needContinue)); + PRF(printf("\nprocessed = %d\n", (unsigned)p->inProcessed)); + + if (res != SZ_OK) + { + PRF(printf("\nWrite error = %d\n", res)); + isErrorMode = True; + p->wasInterrupted = True; + } + if (res != SZ_OK + || (!needContinue && !finish)) + { + PRF(printf("\nWrite Interrupt error = %x\n", res)); + MtDec_Interrupt(p, blockIndex); + } + } + + if (canRecode) + if (!needCode + || res != SZ_OK + || p->wasInterrupted + || codeRes != SZ_OK + || wasInterrupted + || p->numFilledThreads != 0 + || isErrorMode) + { + if (p->numFilledThreads == 0) + p->filledThreadStart = t->index; + if (inDataSize != 0 || !finish) + { + t->inDataSize_Start = inDataSize_Start; + t->inDataSize = inDataSize; + p->numFilledThreads++; + } + PRF(printf("\np->numFilledThreads = %d\n", p->numFilledThreads)); + PRF(printf("p->filledThreadStart = %d\n", p->filledThreadStart)); + } + + if (!finish) + { + RINOK_THREAD(Event_Set(&nextThread->canWrite)); + } + else + { + if (needContinue) + { + // we restore decoding with new iteration + RINOK_THREAD(Event_Set(&p->threads[0].canWrite)); + } + else + { + // we exit from decoding + if (t->index == 0) + return SZ_OK; + p->exitThread = True; + } + RINOK_THREAD(Event_Set(&p->threads[0].canRead)); + } + } + } +} + +#ifdef _WIN32 +#define USE_ALLOCA +#endif + +#ifdef USE_ALLOCA +#ifdef _WIN32 +#include +#else +#include +#endif +#endif + + +static THREAD_FUNC_DECL ThreadFunc1(void *pp) +{ + WRes res; + + CMtDecThread *t = (CMtDecThread *)pp; + CMtDec *p; + + // fprintf(stdout, "\n%d = %p\n", t->index, &t); + + res = ThreadFunc2(t); + p = t->mtDec; + if (res == 0) + return (THREAD_FUNC_RET_TYPE)(UINT_PTR)p->exitThreadWRes; + { + // it's unexpected situation for some threading function error + if (p->exitThreadWRes == 0) + p->exitThreadWRes = res; + PRF(printf("\nthread exit error = %d\n", res)); + p->exitThread = True; + Event_Set(&p->threads[0].canRead); + Event_Set(&p->threads[0].canWrite); + MtProgress_SetError(&p->mtProgress, MY_SRes_HRESULT_FROM_WRes(res)); + } + return (THREAD_FUNC_RET_TYPE)(UINT_PTR)res; +} + +static MY_NO_INLINE THREAD_FUNC_DECL ThreadFunc(void *pp) +{ + #ifdef USE_ALLOCA + CMtDecThread *t = (CMtDecThread *)pp; + // fprintf(stderr, "\n%d = %p - before", t->index, &t); + t->allocaPtr = alloca(t->index * 128); + #endif + return ThreadFunc1(pp); +} + + +int MtDec_PrepareRead(CMtDec *p) +{ + if (p->crossBlock && p->crossStart == p->crossEnd) + { + ISzAlloc_Free(p->alloc, p->crossBlock); + p->crossBlock = NULL; + } + + { + unsigned i; + for (i = 0; i < MTDEC__THREADS_MAX; i++) + if (i > p->numStartedThreads + || p->numFilledThreads <= + (i >= p->filledThreadStart ? + i - p->filledThreadStart : + i + p->numStartedThreads - p->filledThreadStart)) + MtDecThread_FreeInBufs(&p->threads[i]); + } + + return (p->numFilledThreads != 0) || (p->crossStart != p->crossEnd); +} + + +const Byte *MtDec_Read(CMtDec *p, size_t *inLim) +{ + while (p->numFilledThreads != 0) + { + CMtDecThread *t = &p->threads[p->filledThreadStart]; + + if (*inLim != 0) + { + { + void *link = t->inBuf; + void *next = ((CMtDecBufLink *)link)->next; + ISzAlloc_Free(p->alloc, link); + t->inBuf = next; + } + + if (t->inDataSize == 0) + { + MtDecThread_FreeInBufs(t); + if (--p->numFilledThreads == 0) + break; + if (++p->filledThreadStart == p->numStartedThreads) + p->filledThreadStart = 0; + t = &p->threads[p->filledThreadStart]; + } + } + + { + size_t lim = t->inDataSize_Start; + if (lim != 0) + t->inDataSize_Start = 0; + else + { + UInt64 rem = t->inDataSize; + lim = p->inBufSize; + if (lim > rem) + lim = (size_t)rem; + } + t->inDataSize -= lim; + *inLim = lim; + return (const Byte *)MTDEC__DATA_PTR_FROM_LINK(t->inBuf); + } + } + + { + size_t crossSize = p->crossEnd - p->crossStart; + if (crossSize != 0) + { + const Byte *data = MTDEC__DATA_PTR_FROM_LINK(p->crossBlock) + p->crossStart; + *inLim = crossSize; + p->crossStart = 0; + p->crossEnd = 0; + return data; + } + *inLim = 0; + if (p->crossBlock) + { + ISzAlloc_Free(p->alloc, p->crossBlock); + p->crossBlock = NULL; + } + return NULL; + } +} + + +void MtDec_Construct(CMtDec *p) +{ + unsigned i; + + p->inBufSize = (size_t)1 << 18; + + p->numThreadsMax = 0; + + p->inStream = NULL; + + // p->inData = NULL; + // p->inDataSize = 0; + + p->crossBlock = NULL; + p->crossStart = 0; + p->crossEnd = 0; + + p->numFilledThreads = 0; + + p->progress = NULL; + p->alloc = NULL; + + p->mtCallback = NULL; + p->mtCallbackObject = NULL; + + p->allocatedBufsSize = 0; + + for (i = 0; i < MTDEC__THREADS_MAX; i++) + { + CMtDecThread *t = &p->threads[i]; + t->mtDec = p; + t->index = i; + t->inBuf = NULL; + Event_Construct(&t->canRead); + Event_Construct(&t->canWrite); + Thread_Construct(&t->thread); + } + + // Event_Construct(&p->finishedEvent); + + CriticalSection_Init(&p->mtProgress.cs); +} + + +static void MtDec_Free(CMtDec *p) +{ + unsigned i; + + p->exitThread = True; + + for (i = 0; i < MTDEC__THREADS_MAX; i++) + MtDecThread_Destruct(&p->threads[i]); + + // Event_Close(&p->finishedEvent); + + if (p->crossBlock) + { + ISzAlloc_Free(p->alloc, p->crossBlock); + p->crossBlock = NULL; + } +} + + +void MtDec_Destruct(CMtDec *p) +{ + MtDec_Free(p); + + CriticalSection_Delete(&p->mtProgress.cs); +} + + +SRes MtDec_Code(CMtDec *p) +{ + unsigned i; + + p->inProcessed = 0; + + p->blockIndex = 1; // it must be larger than not_defined index (0) + p->isAllocError = False; + p->overflow = False; + p->threadingErrorSRes = SZ_OK; + + p->needContinue = True; + + p->readWasFinished = False; + p->needInterrupt = False; + p->interruptIndex = (UInt64)(Int64)-1; + + p->readProcessed = 0; + p->readRes = SZ_OK; + p->codeRes = SZ_OK; + p->wasInterrupted = False; + + p->crossStart = 0; + p->crossEnd = 0; + + p->filledThreadStart = 0; + p->numFilledThreads = 0; + + { + unsigned numThreads = p->numThreadsMax; + if (numThreads > MTDEC__THREADS_MAX) + numThreads = MTDEC__THREADS_MAX; + p->numStartedThreads_Limit = numThreads; + p->numStartedThreads = 0; + } + + if (p->inBufSize != p->allocatedBufsSize) + { + for (i = 0; i < MTDEC__THREADS_MAX; i++) + { + CMtDecThread *t = &p->threads[i]; + if (t->inBuf) + MtDecThread_FreeInBufs(t); + } + if (p->crossBlock) + { + ISzAlloc_Free(p->alloc, p->crossBlock); + p->crossBlock = NULL; + } + + p->allocatedBufsSize = p->inBufSize; + } + + MtProgress_Init(&p->mtProgress, p->progress); + + // RINOK_THREAD(ArEvent_OptCreate_And_Reset(&p->finishedEvent)); + p->exitThread = False; + p->exitThreadWRes = 0; + + { + WRes wres; + SRes sres; + CMtDecThread *nextThread = &p->threads[p->numStartedThreads++]; + // wres = MtDecThread_CreateAndStart(nextThread); + wres = MtDecThread_CreateEvents(nextThread); + if (wres == 0) { wres = Event_Set(&nextThread->canWrite); + if (wres == 0) { wres = Event_Set(&nextThread->canRead); + if (wres == 0) { THREAD_FUNC_RET_TYPE res = ThreadFunc(nextThread); + wres = (WRes)(UINT_PTR)res; + if (wres != 0) + { + p->needContinue = False; + MtDec_CloseThreads(p); + }}}} + + // wres = 17; // for test + // wres = Event_Wait(&p->finishedEvent); + + sres = MY_SRes_HRESULT_FROM_WRes(wres); + + if (sres != 0) + p->threadingErrorSRes = sres; + + if ( + // wres == 0 + // wres != 0 + // || p->mtc.codeRes == SZ_ERROR_MEM + p->isAllocError + || p->threadingErrorSRes != SZ_OK + || p->overflow) + { + // p->needContinue = True; + } + else + p->needContinue = False; + + if (p->needContinue) + return SZ_OK; + + // if (sres != SZ_OK) + return sres; + // return SZ_ERROR_FAIL; + } +} + +#endif diff --git a/deps/LZMA-SDK/C/MtDec.h b/deps/LZMA-SDK/C/MtDec.h index 7a30b6a9e..c2da46ae2 100644 --- a/deps/LZMA-SDK/C/MtDec.h +++ b/deps/LZMA-SDK/C/MtDec.h @@ -1,202 +1,202 @@ -/* MtDec.h -- Multi-thread Decoder -2020-03-05 : Igor Pavlov : Public domain */ - -#ifndef __MT_DEC_H -#define __MT_DEC_H - -#include "7zTypes.h" - -#ifndef _7ZIP_ST -#include "Threads.h" -#endif - -EXTERN_C_BEGIN - -#ifndef _7ZIP_ST - -#ifndef _7ZIP_ST - #define MTDEC__THREADS_MAX 32 -#else - #define MTDEC__THREADS_MAX 1 -#endif - - -typedef struct -{ - ICompressProgress *progress; - SRes res; - UInt64 totalInSize; - UInt64 totalOutSize; - CCriticalSection cs; -} CMtProgress; - -void MtProgress_Init(CMtProgress *p, ICompressProgress *progress); -SRes MtProgress_Progress_ST(CMtProgress *p); -SRes MtProgress_ProgressAdd(CMtProgress *p, UInt64 inSize, UInt64 outSize); -SRes MtProgress_GetError(CMtProgress *p); -void MtProgress_SetError(CMtProgress *p, SRes res); - -struct _CMtDec; - -typedef struct -{ - struct _CMtDec *mtDec; - unsigned index; - void *inBuf; - - size_t inDataSize_Start; // size of input data in start block - UInt64 inDataSize; // total size of input data in all blocks - - CThread thread; - CAutoResetEvent canRead; - CAutoResetEvent canWrite; - void *allocaPtr; -} CMtDecThread; - -void MtDecThread_FreeInBufs(CMtDecThread *t); - - -typedef enum -{ - MTDEC_PARSE_CONTINUE, // continue this block with more input data - MTDEC_PARSE_OVERFLOW, // MT buffers overflow, need switch to single-thread - MTDEC_PARSE_NEW, // new block - MTDEC_PARSE_END // end of block threading. But we still can return to threading after Write(&needContinue) -} EMtDecParseState; - -typedef struct -{ - // in - int startCall; - const Byte *src; - size_t srcSize; - // in : (srcSize == 0) is allowed - // out : it's allowed to return less that actually was used ? - int srcFinished; - - // out - EMtDecParseState state; - BoolInt canCreateNewThread; - UInt64 outPos; // check it (size_t) -} CMtDecCallbackInfo; - - -typedef struct -{ - void (*Parse)(void *p, unsigned coderIndex, CMtDecCallbackInfo *ci); - - // PreCode() and Code(): - // (SRes_return_result != SZ_OK) means stop decoding, no need another blocks - SRes (*PreCode)(void *p, unsigned coderIndex); - SRes (*Code)(void *p, unsigned coderIndex, - const Byte *src, size_t srcSize, int srcFinished, - UInt64 *inCodePos, UInt64 *outCodePos, int *stop); - // stop - means stop another Code calls - - - /* Write() must be called, if Parse() was called - set (needWrite) if - { - && (was not interrupted by progress) - && (was not interrupted in previous block) - } - - out: - if (*needContinue), decoder still need to continue decoding with new iteration, - even after MTDEC_PARSE_END - if (*canRecode), we didn't flush current block data, so we still can decode current block later. - */ - SRes (*Write)(void *p, unsigned coderIndex, - BoolInt needWriteToStream, - const Byte *src, size_t srcSize, BoolInt isCross, - // int srcFinished, - BoolInt *needContinue, - BoolInt *canRecode); - -} IMtDecCallback2; - - - -typedef struct _CMtDec -{ - /* input variables */ - - size_t inBufSize; /* size of input block */ - unsigned numThreadsMax; - // size_t inBlockMax; - unsigned numThreadsMax_2; - - ISeqInStream *inStream; - // const Byte *inData; - // size_t inDataSize; - - ICompressProgress *progress; - ISzAllocPtr alloc; - - IMtDecCallback2 *mtCallback; - void *mtCallbackObject; - - - /* internal variables */ - - size_t allocatedBufsSize; - - BoolInt exitThread; - WRes exitThreadWRes; - - UInt64 blockIndex; - BoolInt isAllocError; - BoolInt overflow; - SRes threadingErrorSRes; - - BoolInt needContinue; - - // CAutoResetEvent finishedEvent; - - SRes readRes; - SRes codeRes; - - BoolInt wasInterrupted; - - unsigned numStartedThreads_Limit; - unsigned numStartedThreads; - - Byte *crossBlock; - size_t crossStart; - size_t crossEnd; - UInt64 readProcessed; - BoolInt readWasFinished; - UInt64 inProcessed; - - unsigned filledThreadStart; - unsigned numFilledThreads; - - #ifndef _7ZIP_ST - BoolInt needInterrupt; - UInt64 interruptIndex; - CMtProgress mtProgress; - CMtDecThread threads[MTDEC__THREADS_MAX]; - #endif -} CMtDec; - - -void MtDec_Construct(CMtDec *p); -void MtDec_Destruct(CMtDec *p); - -/* -MtDec_Code() returns: - SZ_OK - in most cases - MY_SRes_HRESULT_FROM_WRes(WRes_error) - in case of unexpected error in threading function -*/ - -SRes MtDec_Code(CMtDec *p); -Byte *MtDec_GetCrossBuff(CMtDec *p); - -int MtDec_PrepareRead(CMtDec *p); -const Byte *MtDec_Read(CMtDec *p, size_t *inLim); - -#endif - -EXTERN_C_END - -#endif +/* MtDec.h -- Multi-thread Decoder +2020-03-05 : Igor Pavlov : Public domain */ + +#ifndef __MT_DEC_H +#define __MT_DEC_H + +#include "7zTypes.h" + +#ifndef _7ZIP_ST +#include "Threads.h" +#endif + +EXTERN_C_BEGIN + +#ifndef _7ZIP_ST + +#ifndef _7ZIP_ST + #define MTDEC__THREADS_MAX 32 +#else + #define MTDEC__THREADS_MAX 1 +#endif + + +typedef struct +{ + ICompressProgress *progress; + SRes res; + UInt64 totalInSize; + UInt64 totalOutSize; + CCriticalSection cs; +} CMtProgress; + +void MtProgress_Init(CMtProgress *p, ICompressProgress *progress); +SRes MtProgress_Progress_ST(CMtProgress *p); +SRes MtProgress_ProgressAdd(CMtProgress *p, UInt64 inSize, UInt64 outSize); +SRes MtProgress_GetError(CMtProgress *p); +void MtProgress_SetError(CMtProgress *p, SRes res); + +struct _CMtDec; + +typedef struct +{ + struct _CMtDec *mtDec; + unsigned index; + void *inBuf; + + size_t inDataSize_Start; // size of input data in start block + UInt64 inDataSize; // total size of input data in all blocks + + CThread thread; + CAutoResetEvent canRead; + CAutoResetEvent canWrite; + void *allocaPtr; +} CMtDecThread; + +void MtDecThread_FreeInBufs(CMtDecThread *t); + + +typedef enum +{ + MTDEC_PARSE_CONTINUE, // continue this block with more input data + MTDEC_PARSE_OVERFLOW, // MT buffers overflow, need switch to single-thread + MTDEC_PARSE_NEW, // new block + MTDEC_PARSE_END // end of block threading. But we still can return to threading after Write(&needContinue) +} EMtDecParseState; + +typedef struct +{ + // in + int startCall; + const Byte *src; + size_t srcSize; + // in : (srcSize == 0) is allowed + // out : it's allowed to return less that actually was used ? + int srcFinished; + + // out + EMtDecParseState state; + BoolInt canCreateNewThread; + UInt64 outPos; // check it (size_t) +} CMtDecCallbackInfo; + + +typedef struct +{ + void (*Parse)(void *p, unsigned coderIndex, CMtDecCallbackInfo *ci); + + // PreCode() and Code(): + // (SRes_return_result != SZ_OK) means stop decoding, no need another blocks + SRes (*PreCode)(void *p, unsigned coderIndex); + SRes (*Code)(void *p, unsigned coderIndex, + const Byte *src, size_t srcSize, int srcFinished, + UInt64 *inCodePos, UInt64 *outCodePos, int *stop); + // stop - means stop another Code calls + + + /* Write() must be called, if Parse() was called + set (needWrite) if + { + && (was not interrupted by progress) + && (was not interrupted in previous block) + } + + out: + if (*needContinue), decoder still need to continue decoding with new iteration, + even after MTDEC_PARSE_END + if (*canRecode), we didn't flush current block data, so we still can decode current block later. + */ + SRes (*Write)(void *p, unsigned coderIndex, + BoolInt needWriteToStream, + const Byte *src, size_t srcSize, BoolInt isCross, + // int srcFinished, + BoolInt *needContinue, + BoolInt *canRecode); + +} IMtDecCallback2; + + + +typedef struct _CMtDec +{ + /* input variables */ + + size_t inBufSize; /* size of input block */ + unsigned numThreadsMax; + // size_t inBlockMax; + unsigned numThreadsMax_2; + + ISeqInStream *inStream; + // const Byte *inData; + // size_t inDataSize; + + ICompressProgress *progress; + ISzAllocPtr alloc; + + IMtDecCallback2 *mtCallback; + void *mtCallbackObject; + + + /* internal variables */ + + size_t allocatedBufsSize; + + BoolInt exitThread; + WRes exitThreadWRes; + + UInt64 blockIndex; + BoolInt isAllocError; + BoolInt overflow; + SRes threadingErrorSRes; + + BoolInt needContinue; + + // CAutoResetEvent finishedEvent; + + SRes readRes; + SRes codeRes; + + BoolInt wasInterrupted; + + unsigned numStartedThreads_Limit; + unsigned numStartedThreads; + + Byte *crossBlock; + size_t crossStart; + size_t crossEnd; + UInt64 readProcessed; + BoolInt readWasFinished; + UInt64 inProcessed; + + unsigned filledThreadStart; + unsigned numFilledThreads; + + #ifndef _7ZIP_ST + BoolInt needInterrupt; + UInt64 interruptIndex; + CMtProgress mtProgress; + CMtDecThread threads[MTDEC__THREADS_MAX]; + #endif +} CMtDec; + + +void MtDec_Construct(CMtDec *p); +void MtDec_Destruct(CMtDec *p); + +/* +MtDec_Code() returns: + SZ_OK - in most cases + MY_SRes_HRESULT_FROM_WRes(WRes_error) - in case of unexpected error in threading function +*/ + +SRes MtDec_Code(CMtDec *p); +Byte *MtDec_GetCrossBuff(CMtDec *p); + +int MtDec_PrepareRead(CMtDec *p); +const Byte *MtDec_Read(CMtDec *p, size_t *inLim); + +#endif + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/Ppmd.h b/deps/LZMA-SDK/C/Ppmd.h index ee93ecece..b19879208 100644 --- a/deps/LZMA-SDK/C/Ppmd.h +++ b/deps/LZMA-SDK/C/Ppmd.h @@ -1,167 +1,167 @@ -/* Ppmd.h -- PPMD codec common code -2021-04-13 : Igor Pavlov : Public domain -This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */ - -#ifndef __PPMD_H -#define __PPMD_H - -#include "CpuArch.h" - -EXTERN_C_BEGIN - -#if defined(MY_CPU_SIZEOF_POINTER) && (MY_CPU_SIZEOF_POINTER == 4) -/* - PPMD code always uses 32-bit internal fields in PPMD structures to store internal references in main block. - if (PPMD_32BIT is defined), the PPMD code stores internal pointers to 32-bit reference fields. - if (PPMD_32BIT is NOT defined), the PPMD code stores internal UInt32 offsets to reference fields. - if (pointer size is 64-bit), then (PPMD_32BIT) mode is not allowed, - if (pointer size is 32-bit), then (PPMD_32BIT) mode is optional, - and it's allowed to disable PPMD_32BIT mode even if pointer is 32-bit. - PPMD code works slightly faster in (PPMD_32BIT) mode. -*/ - #define PPMD_32BIT -#endif - -#define PPMD_INT_BITS 7 -#define PPMD_PERIOD_BITS 7 -#define PPMD_BIN_SCALE (1 << (PPMD_INT_BITS + PPMD_PERIOD_BITS)) - -#define PPMD_GET_MEAN_SPEC(summ, shift, round) (((summ) + (1 << ((shift) - (round)))) >> (shift)) -#define PPMD_GET_MEAN(summ) PPMD_GET_MEAN_SPEC((summ), PPMD_PERIOD_BITS, 2) -#define PPMD_UPDATE_PROB_0(prob) ((prob) + (1 << PPMD_INT_BITS) - PPMD_GET_MEAN(prob)) -#define PPMD_UPDATE_PROB_1(prob) ((prob) - PPMD_GET_MEAN(prob)) - -#define PPMD_N1 4 -#define PPMD_N2 4 -#define PPMD_N3 4 -#define PPMD_N4 ((128 + 3 - 1 * PPMD_N1 - 2 * PPMD_N2 - 3 * PPMD_N3) / 4) -#define PPMD_NUM_INDEXES (PPMD_N1 + PPMD_N2 + PPMD_N3 + PPMD_N4) - -MY_CPU_pragma_pack_push_1 -/* Most compilers works OK here even without #pragma pack(push, 1), but some GCC compilers need it. */ - -/* SEE-contexts for PPM-contexts with masked symbols */ -typedef struct -{ - UInt16 Summ; /* Freq */ - Byte Shift; /* Speed of Freq change; low Shift is for fast change */ - Byte Count; /* Count to next change of Shift */ -} CPpmd_See; - -#define Ppmd_See_Update(p) if ((p)->Shift < PPMD_PERIOD_BITS && --(p)->Count == 0) \ - { (p)->Summ = (UInt16)((p)->Summ << 1); (p)->Count = (Byte)(3 << (p)->Shift++); } - - -typedef struct -{ - Byte Symbol; - Byte Freq; - UInt16 Successor_0; - UInt16 Successor_1; -} CPpmd_State; - -typedef struct CPpmd_State2_ -{ - Byte Symbol; - Byte Freq; -} CPpmd_State2; - -typedef struct CPpmd_State4_ -{ - UInt16 Successor_0; - UInt16 Successor_1; -} CPpmd_State4; - -MY_CPU_pragma_pop - -/* - PPMD code can write full CPpmd_State structure data to CPpmd*_Context - at (byte offset = 2) instead of some fields of original CPpmd*_Context structure. - - If we use pointers to different types, but that point to shared - memory space, we can have aliasing problem (strict aliasing). - - XLC compiler in -O2 mode can change the order of memory write instructions - in relation to read instructions, if we have use pointers to different types. - - To solve that aliasing problem we use combined CPpmd*_Context structure - with unions that contain the fields from both structures: - the original CPpmd*_Context and CPpmd_State. - So we can access the fields from both structures via one pointer, - and the compiler doesn't change the order of write instructions - in relation to read instructions. - - If we don't use memory write instructions to shared memory in - some local code, and we use only reading instructions (read only), - then probably it's safe to use pointers to different types for reading. -*/ - - - -#ifdef PPMD_32BIT - - #define Ppmd_Ref_Type(type) type * - #define Ppmd_GetRef(p, ptr) (ptr) - #define Ppmd_GetPtr(p, ptr) (ptr) - #define Ppmd_GetPtr_Type(p, ptr, note_type) (ptr) - -#else - - #define Ppmd_Ref_Type(type) UInt32 - #define Ppmd_GetRef(p, ptr) ((UInt32)((Byte *)(ptr) - (p)->Base)) - #define Ppmd_GetPtr(p, offs) ((void *)((p)->Base + (offs))) - #define Ppmd_GetPtr_Type(p, offs, type) ((type *)Ppmd_GetPtr(p, offs)) - -#endif // PPMD_32BIT - - -typedef Ppmd_Ref_Type(CPpmd_State) CPpmd_State_Ref; -typedef Ppmd_Ref_Type(void) CPpmd_Void_Ref; -typedef Ppmd_Ref_Type(Byte) CPpmd_Byte_Ref; - - -/* -#ifdef MY_CPU_LE_UNALIGN -// the unaligned 32-bit access latency can be too large, if the data is not in L1 cache. -#define Ppmd_GET_SUCCESSOR(p) ((CPpmd_Void_Ref)*(const UInt32 *)(const void *)&(p)->Successor_0) -#define Ppmd_SET_SUCCESSOR(p, v) *(UInt32 *)(void *)(void *)&(p)->Successor_0 = (UInt32)(v) - -#else -*/ - -/* - We can write 16-bit halves to 32-bit (Successor) field in any selected order. - But the native order is more consistent way. - So we use the native order, if LE/BE order can be detected here at compile time. -*/ - -#ifdef MY_CPU_BE - - #define Ppmd_GET_SUCCESSOR(p) \ - ( (CPpmd_Void_Ref) (((UInt32)(p)->Successor_0 << 16) | (p)->Successor_1) ) - - #define Ppmd_SET_SUCCESSOR(p, v) { \ - (p)->Successor_0 = (UInt16)(((UInt32)(v) >> 16) /* & 0xFFFF */); \ - (p)->Successor_1 = (UInt16)((UInt32)(v) /* & 0xFFFF */); } - -#else - - #define Ppmd_GET_SUCCESSOR(p) \ - ( (CPpmd_Void_Ref) ((p)->Successor_0 | ((UInt32)(p)->Successor_1 << 16)) ) - - #define Ppmd_SET_SUCCESSOR(p, v) { \ - (p)->Successor_0 = (UInt16)((UInt32)(v) /* & 0xFFFF */); \ - (p)->Successor_1 = (UInt16)(((UInt32)(v) >> 16) /* & 0xFFFF */); } - -#endif - -// #endif - - -#define PPMD_SetAllBitsIn256Bytes(p) \ - { size_t z; for (z = 0; z < 256 / sizeof(p[0]); z += 8) { \ - p[z+7] = p[z+6] = p[z+5] = p[z+4] = p[z+3] = p[z+2] = p[z+1] = p[z+0] = ~(size_t)0; }} - -EXTERN_C_END - -#endif +/* Ppmd.h -- PPMD codec common code +2021-04-13 : Igor Pavlov : Public domain +This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */ + +#ifndef __PPMD_H +#define __PPMD_H + +#include "CpuArch.h" + +EXTERN_C_BEGIN + +#if defined(MY_CPU_SIZEOF_POINTER) && (MY_CPU_SIZEOF_POINTER == 4) +/* + PPMD code always uses 32-bit internal fields in PPMD structures to store internal references in main block. + if (PPMD_32BIT is defined), the PPMD code stores internal pointers to 32-bit reference fields. + if (PPMD_32BIT is NOT defined), the PPMD code stores internal UInt32 offsets to reference fields. + if (pointer size is 64-bit), then (PPMD_32BIT) mode is not allowed, + if (pointer size is 32-bit), then (PPMD_32BIT) mode is optional, + and it's allowed to disable PPMD_32BIT mode even if pointer is 32-bit. + PPMD code works slightly faster in (PPMD_32BIT) mode. +*/ + #define PPMD_32BIT +#endif + +#define PPMD_INT_BITS 7 +#define PPMD_PERIOD_BITS 7 +#define PPMD_BIN_SCALE (1 << (PPMD_INT_BITS + PPMD_PERIOD_BITS)) + +#define PPMD_GET_MEAN_SPEC(summ, shift, round) (((summ) + (1 << ((shift) - (round)))) >> (shift)) +#define PPMD_GET_MEAN(summ) PPMD_GET_MEAN_SPEC((summ), PPMD_PERIOD_BITS, 2) +#define PPMD_UPDATE_PROB_0(prob) ((prob) + (1 << PPMD_INT_BITS) - PPMD_GET_MEAN(prob)) +#define PPMD_UPDATE_PROB_1(prob) ((prob) - PPMD_GET_MEAN(prob)) + +#define PPMD_N1 4 +#define PPMD_N2 4 +#define PPMD_N3 4 +#define PPMD_N4 ((128 + 3 - 1 * PPMD_N1 - 2 * PPMD_N2 - 3 * PPMD_N3) / 4) +#define PPMD_NUM_INDEXES (PPMD_N1 + PPMD_N2 + PPMD_N3 + PPMD_N4) + +MY_CPU_pragma_pack_push_1 +/* Most compilers works OK here even without #pragma pack(push, 1), but some GCC compilers need it. */ + +/* SEE-contexts for PPM-contexts with masked symbols */ +typedef struct +{ + UInt16 Summ; /* Freq */ + Byte Shift; /* Speed of Freq change; low Shift is for fast change */ + Byte Count; /* Count to next change of Shift */ +} CPpmd_See; + +#define Ppmd_See_Update(p) if ((p)->Shift < PPMD_PERIOD_BITS && --(p)->Count == 0) \ + { (p)->Summ = (UInt16)((p)->Summ << 1); (p)->Count = (Byte)(3 << (p)->Shift++); } + + +typedef struct +{ + Byte Symbol; + Byte Freq; + UInt16 Successor_0; + UInt16 Successor_1; +} CPpmd_State; + +typedef struct CPpmd_State2_ +{ + Byte Symbol; + Byte Freq; +} CPpmd_State2; + +typedef struct CPpmd_State4_ +{ + UInt16 Successor_0; + UInt16 Successor_1; +} CPpmd_State4; + +MY_CPU_pragma_pop + +/* + PPMD code can write full CPpmd_State structure data to CPpmd*_Context + at (byte offset = 2) instead of some fields of original CPpmd*_Context structure. + + If we use pointers to different types, but that point to shared + memory space, we can have aliasing problem (strict aliasing). + + XLC compiler in -O2 mode can change the order of memory write instructions + in relation to read instructions, if we have use pointers to different types. + + To solve that aliasing problem we use combined CPpmd*_Context structure + with unions that contain the fields from both structures: + the original CPpmd*_Context and CPpmd_State. + So we can access the fields from both structures via one pointer, + and the compiler doesn't change the order of write instructions + in relation to read instructions. + + If we don't use memory write instructions to shared memory in + some local code, and we use only reading instructions (read only), + then probably it's safe to use pointers to different types for reading. +*/ + + + +#ifdef PPMD_32BIT + + #define Ppmd_Ref_Type(type) type * + #define Ppmd_GetRef(p, ptr) (ptr) + #define Ppmd_GetPtr(p, ptr) (ptr) + #define Ppmd_GetPtr_Type(p, ptr, note_type) (ptr) + +#else + + #define Ppmd_Ref_Type(type) UInt32 + #define Ppmd_GetRef(p, ptr) ((UInt32)((Byte *)(ptr) - (p)->Base)) + #define Ppmd_GetPtr(p, offs) ((void *)((p)->Base + (offs))) + #define Ppmd_GetPtr_Type(p, offs, type) ((type *)Ppmd_GetPtr(p, offs)) + +#endif // PPMD_32BIT + + +typedef Ppmd_Ref_Type(CPpmd_State) CPpmd_State_Ref; +typedef Ppmd_Ref_Type(void) CPpmd_Void_Ref; +typedef Ppmd_Ref_Type(Byte) CPpmd_Byte_Ref; + + +/* +#ifdef MY_CPU_LE_UNALIGN +// the unaligned 32-bit access latency can be too large, if the data is not in L1 cache. +#define Ppmd_GET_SUCCESSOR(p) ((CPpmd_Void_Ref)*(const UInt32 *)(const void *)&(p)->Successor_0) +#define Ppmd_SET_SUCCESSOR(p, v) *(UInt32 *)(void *)(void *)&(p)->Successor_0 = (UInt32)(v) + +#else +*/ + +/* + We can write 16-bit halves to 32-bit (Successor) field in any selected order. + But the native order is more consistent way. + So we use the native order, if LE/BE order can be detected here at compile time. +*/ + +#ifdef MY_CPU_BE + + #define Ppmd_GET_SUCCESSOR(p) \ + ( (CPpmd_Void_Ref) (((UInt32)(p)->Successor_0 << 16) | (p)->Successor_1) ) + + #define Ppmd_SET_SUCCESSOR(p, v) { \ + (p)->Successor_0 = (UInt16)(((UInt32)(v) >> 16) /* & 0xFFFF */); \ + (p)->Successor_1 = (UInt16)((UInt32)(v) /* & 0xFFFF */); } + +#else + + #define Ppmd_GET_SUCCESSOR(p) \ + ( (CPpmd_Void_Ref) ((p)->Successor_0 | ((UInt32)(p)->Successor_1 << 16)) ) + + #define Ppmd_SET_SUCCESSOR(p, v) { \ + (p)->Successor_0 = (UInt16)((UInt32)(v) /* & 0xFFFF */); \ + (p)->Successor_1 = (UInt16)(((UInt32)(v) >> 16) /* & 0xFFFF */); } + +#endif + +// #endif + + +#define PPMD_SetAllBitsIn256Bytes(p) \ + { size_t z; for (z = 0; z < 256 / sizeof(p[0]); z += 8) { \ + p[z+7] = p[z+6] = p[z+5] = p[z+4] = p[z+3] = p[z+2] = p[z+1] = p[z+0] = ~(size_t)0; }} + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/Ppmd7.c b/deps/LZMA-SDK/C/Ppmd7.c index b6ecf1430..cf401cb37 100644 --- a/deps/LZMA-SDK/C/Ppmd7.c +++ b/deps/LZMA-SDK/C/Ppmd7.c @@ -1,1104 +1,1104 @@ -/* Ppmd7.c -- PPMdH codec -2021-04-13 : Igor Pavlov : Public domain -This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */ - -#include "Precomp.h" - -#include - -#include "Ppmd7.h" - -/* define PPMD7_ORDER_0_SUPPPORT to suport order-0 mode, unsupported by orignal PPMd var.H. code */ -// #define PPMD7_ORDER_0_SUPPPORT - -MY_ALIGN(16) -static const Byte PPMD7_kExpEscape[16] = { 25, 14, 9, 7, 5, 5, 4, 4, 4, 3, 3, 3, 2, 2, 2, 2 }; -MY_ALIGN(16) -static const UInt16 kInitBinEsc[] = { 0x3CDD, 0x1F3F, 0x59BF, 0x48F3, 0x64A1, 0x5ABC, 0x6632, 0x6051}; - -#define MAX_FREQ 124 -#define UNIT_SIZE 12 - -#define U2B(nu) ((UInt32)(nu) * UNIT_SIZE) -#define U2I(nu) (p->Units2Indx[(size_t)(nu) - 1]) -#define I2U(indx) ((unsigned)p->Indx2Units[indx]) -#define I2U_UInt16(indx) ((UInt16)p->Indx2Units[indx]) - -#define REF(ptr) Ppmd_GetRef(p, ptr) - -#define STATS_REF(ptr) ((CPpmd_State_Ref)REF(ptr)) - -#define CTX(ref) ((CPpmd7_Context *)Ppmd7_GetContext(p, ref)) -#define STATS(ctx) Ppmd7_GetStats(p, ctx) -#define ONE_STATE(ctx) Ppmd7Context_OneState(ctx) -#define SUFFIX(ctx) CTX((ctx)->Suffix) - -typedef CPpmd7_Context * CTX_PTR; - -struct CPpmd7_Node_; - -typedef Ppmd_Ref_Type(struct CPpmd7_Node_) CPpmd7_Node_Ref; - -typedef struct CPpmd7_Node_ -{ - UInt16 Stamp; /* must be at offset 0 as CPpmd7_Context::NumStats. Stamp=0 means free */ - UInt16 NU; - CPpmd7_Node_Ref Next; /* must be at offset >= 4 */ - CPpmd7_Node_Ref Prev; -} CPpmd7_Node; - -#define NODE(r) Ppmd_GetPtr_Type(p, r, CPpmd7_Node) - -void Ppmd7_Construct(CPpmd7 *p) -{ - unsigned i, k, m; - - p->Base = NULL; - - for (i = 0, k = 0; i < PPMD_NUM_INDEXES; i++) - { - unsigned step = (i >= 12 ? 4 : (i >> 2) + 1); - do { p->Units2Indx[k++] = (Byte)i; } while (--step); - p->Indx2Units[i] = (Byte)k; - } - - p->NS2BSIndx[0] = (0 << 1); - p->NS2BSIndx[1] = (1 << 1); - memset(p->NS2BSIndx + 2, (2 << 1), 9); - memset(p->NS2BSIndx + 11, (3 << 1), 256 - 11); - - for (i = 0; i < 3; i++) - p->NS2Indx[i] = (Byte)i; - - for (m = i, k = 1; i < 256; i++) - { - p->NS2Indx[i] = (Byte)m; - if (--k == 0) - k = (++m) - 2; - } - - memcpy(p->ExpEscape, PPMD7_kExpEscape, 16); -} - - -void Ppmd7_Free(CPpmd7 *p, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, p->Base); - p->Size = 0; - p->Base = NULL; -} - - -BoolInt Ppmd7_Alloc(CPpmd7 *p, UInt32 size, ISzAllocPtr alloc) -{ - if (!p->Base || p->Size != size) - { - Ppmd7_Free(p, alloc); - p->AlignOffset = (4 - size) & 3; - if ((p->Base = (Byte *)ISzAlloc_Alloc(alloc, p->AlignOffset + size)) == NULL) - return False; - p->Size = size; - } - return True; -} - - - -// ---------- Internal Memory Allocator ---------- - -/* We can use CPpmd7_Node in list of free units (as in Ppmd8) - But we still need one additional list walk pass in GlueFreeBlocks(). - So we use simple CPpmd_Void_Ref instead of CPpmd7_Node in InsertNode() / RemoveNode() -*/ - -#define EMPTY_NODE 0 - - -static void InsertNode(CPpmd7 *p, void *node, unsigned indx) -{ - *((CPpmd_Void_Ref *)node) = p->FreeList[indx]; - // ((CPpmd7_Node *)node)->Next = (CPpmd7_Node_Ref)p->FreeList[indx]; - - p->FreeList[indx] = REF(node); - -} - - -static void *RemoveNode(CPpmd7 *p, unsigned indx) -{ - CPpmd_Void_Ref *node = (CPpmd_Void_Ref *)Ppmd7_GetPtr(p, p->FreeList[indx]); - p->FreeList[indx] = *node; - // CPpmd7_Node *node = NODE((CPpmd7_Node_Ref)p->FreeList[indx]); - // p->FreeList[indx] = node->Next; - return node; -} - - -static void SplitBlock(CPpmd7 *p, void *ptr, unsigned oldIndx, unsigned newIndx) -{ - unsigned i, nu = I2U(oldIndx) - I2U(newIndx); - ptr = (Byte *)ptr + U2B(I2U(newIndx)); - if (I2U(i = U2I(nu)) != nu) - { - unsigned k = I2U(--i); - InsertNode(p, ((Byte *)ptr) + U2B(k), nu - k - 1); - } - InsertNode(p, ptr, i); -} - - -/* we use CPpmd7_Node_Union union to solve XLC -O2 strict pointer aliasing problem */ - -typedef union _CPpmd7_Node_Union -{ - CPpmd7_Node Node; - CPpmd7_Node_Ref NextRef; -} CPpmd7_Node_Union; - -/* Original PPmdH (Ppmd7) code uses doubly linked list in GlueFreeBlocks() - we use single linked list similar to Ppmd8 code */ - - -static void GlueFreeBlocks(CPpmd7 *p) -{ - /* - we use first UInt16 field of 12-bytes UNITs as record type stamp - CPpmd_State { Byte Symbol; Byte Freq; : Freq != 0 - CPpmd7_Context { UInt16 NumStats; : NumStats != 0 - CPpmd7_Node { UInt16 Stamp : Stamp == 0 for free record - : Stamp == 1 for head record and guard - Last 12-bytes UNIT in array is always contains 12-bytes order-0 CPpmd7_Context record. - */ - CPpmd7_Node_Ref head, n = 0; - - p->GlueCount = 255; - - - /* we set guard NODE at LoUnit */ - if (p->LoUnit != p->HiUnit) - ((CPpmd7_Node *)(void *)p->LoUnit)->Stamp = 1; - - { - /* Create list of free blocks. - We still need one additional list walk pass before Glue. */ - unsigned i; - for (i = 0; i < PPMD_NUM_INDEXES; i++) - { - const UInt16 nu = I2U_UInt16(i); - CPpmd7_Node_Ref next = (CPpmd7_Node_Ref)p->FreeList[i]; - p->FreeList[i] = 0; - while (next != 0) - { - /* Don't change the order of the following commands: */ - CPpmd7_Node_Union *un = (CPpmd7_Node_Union *)NODE(next); - const CPpmd7_Node_Ref tmp = next; - next = un->NextRef; - un->Node.Stamp = EMPTY_NODE; - un->Node.NU = nu; - un->Node.Next = n; - n = tmp; - } - } - } - - head = n; - /* Glue and Fill must walk the list in same direction */ - { - /* Glue free blocks */ - CPpmd7_Node_Ref *prev = &head; - while (n) - { - CPpmd7_Node *node = NODE(n); - UInt32 nu = node->NU; - n = node->Next; - if (nu == 0) - { - *prev = n; - continue; - } - prev = &node->Next; - for (;;) - { - CPpmd7_Node *node2 = node + nu; - nu += node2->NU; - if (node2->Stamp != EMPTY_NODE || nu >= 0x10000) - break; - node->NU = (UInt16)nu; - node2->NU = 0; - } - } - } - - /* Fill lists of free blocks */ - for (n = head; n != 0;) - { - CPpmd7_Node *node = NODE(n); - UInt32 nu = node->NU; - unsigned i; - n = node->Next; - if (nu == 0) - continue; - for (; nu > 128; nu -= 128, node += 128) - InsertNode(p, node, PPMD_NUM_INDEXES - 1); - if (I2U(i = U2I(nu)) != nu) - { - unsigned k = I2U(--i); - InsertNode(p, node + k, (unsigned)nu - k - 1); - } - InsertNode(p, node, i); - } -} - - -MY_NO_INLINE -static void *AllocUnitsRare(CPpmd7 *p, unsigned indx) -{ - unsigned i; - - if (p->GlueCount == 0) - { - GlueFreeBlocks(p); - if (p->FreeList[indx] != 0) - return RemoveNode(p, indx); - } - - i = indx; - - do - { - if (++i == PPMD_NUM_INDEXES) - { - UInt32 numBytes = U2B(I2U(indx)); - Byte *us = p->UnitsStart; - p->GlueCount--; - return ((UInt32)(us - p->Text) > numBytes) ? (p->UnitsStart = us - numBytes) : NULL; - } - } - while (p->FreeList[i] == 0); - - { - void *block = RemoveNode(p, i); - SplitBlock(p, block, i, indx); - return block; - } -} - - -static void *AllocUnits(CPpmd7 *p, unsigned indx) -{ - if (p->FreeList[indx] != 0) - return RemoveNode(p, indx); - { - UInt32 numBytes = U2B(I2U(indx)); - Byte *lo = p->LoUnit; - if ((UInt32)(p->HiUnit - lo) >= numBytes) - { - p->LoUnit = lo + numBytes; - return lo; - } - } - return AllocUnitsRare(p, indx); -} - - -#define MyMem12Cpy(dest, src, num) \ - { UInt32 *d = (UInt32 *)dest; const UInt32 *z = (const UInt32 *)src; UInt32 n = num; \ - do { d[0] = z[0]; d[1] = z[1]; d[2] = z[2]; z += 3; d += 3; } while (--n); } - - -/* -static void *ShrinkUnits(CPpmd7 *p, void *oldPtr, unsigned oldNU, unsigned newNU) -{ - unsigned i0 = U2I(oldNU); - unsigned i1 = U2I(newNU); - if (i0 == i1) - return oldPtr; - if (p->FreeList[i1] != 0) - { - void *ptr = RemoveNode(p, i1); - MyMem12Cpy(ptr, oldPtr, newNU); - InsertNode(p, oldPtr, i0); - return ptr; - } - SplitBlock(p, oldPtr, i0, i1); - return oldPtr; -} -*/ - - -#define SUCCESSOR(p) Ppmd_GET_SUCCESSOR(p) -static void SetSuccessor(CPpmd_State *p, CPpmd_Void_Ref v) -{ - Ppmd_SET_SUCCESSOR(p, v); -} - - - -MY_NO_INLINE -static -void RestartModel(CPpmd7 *p) -{ - unsigned i, k; - - memset(p->FreeList, 0, sizeof(p->FreeList)); - - p->Text = p->Base + p->AlignOffset; - p->HiUnit = p->Text + p->Size; - p->LoUnit = p->UnitsStart = p->HiUnit - p->Size / 8 / UNIT_SIZE * 7 * UNIT_SIZE; - p->GlueCount = 0; - - p->OrderFall = p->MaxOrder; - p->RunLength = p->InitRL = -(Int32)((p->MaxOrder < 12) ? p->MaxOrder : 12) - 1; - p->PrevSuccess = 0; - - { - CPpmd7_Context *mc = (CTX_PTR)(void *)(p->HiUnit -= UNIT_SIZE); /* AllocContext(p); */ - CPpmd_State *s = (CPpmd_State *)p->LoUnit; /* AllocUnits(p, PPMD_NUM_INDEXES - 1); */ - - p->LoUnit += U2B(256 / 2); - p->MaxContext = p->MinContext = mc; - p->FoundState = s; - - mc->NumStats = 256; - mc->Union2.SummFreq = 256 + 1; - mc->Union4.Stats = REF(s); - mc->Suffix = 0; - - for (i = 0; i < 256; i++, s++) - { - s->Symbol = (Byte)i; - s->Freq = 1; - SetSuccessor(s, 0); - } - - #ifdef PPMD7_ORDER_0_SUPPPORT - if (p->MaxOrder == 0) - { - CPpmd_Void_Ref r = REF(mc); - s = p->FoundState; - for (i = 0; i < 256; i++, s++) - SetSuccessor(s, r); - return; - } - #endif - } - - for (i = 0; i < 128; i++) - - - - for (k = 0; k < 8; k++) - { - unsigned m; - UInt16 *dest = p->BinSumm[i] + k; - UInt16 val = (UInt16)(PPMD_BIN_SCALE - kInitBinEsc[k] / (i + 2)); - for (m = 0; m < 64; m += 8) - dest[m] = val; - } - - - for (i = 0; i < 25; i++) - { - - CPpmd_See *s = p->See[i]; - - - - unsigned summ = ((5 * i + 10) << (PPMD_PERIOD_BITS - 4)); - for (k = 0; k < 16; k++, s++) - { - s->Summ = (UInt16)summ; - s->Shift = (PPMD_PERIOD_BITS - 4); - s->Count = 4; - } - } - - p->DummySee.Summ = 0; /* unused */ - p->DummySee.Shift = PPMD_PERIOD_BITS; - p->DummySee.Count = 64; /* unused */ -} - - -void Ppmd7_Init(CPpmd7 *p, unsigned maxOrder) -{ - p->MaxOrder = maxOrder; - - RestartModel(p); -} - - - -/* - CreateSuccessors() - It's called when (FoundState->Successor) is RAW-Successor, - that is the link to position in Raw text. - So we create Context records and write the links to - FoundState->Successor and to identical RAW-Successors in suffix - contexts of MinContex. - - The function returns: - if (OrderFall == 0) then MinContext is already at MAX order, - { return pointer to new or existing context of same MAX order } - else - { return pointer to new real context that will be (Order+1) in comparison with MinContext - - also it can return pointer to real context of same order, -*/ - -MY_NO_INLINE -static CTX_PTR CreateSuccessors(CPpmd7 *p) -{ - CTX_PTR c = p->MinContext; - CPpmd_Byte_Ref upBranch = (CPpmd_Byte_Ref)SUCCESSOR(p->FoundState); - Byte newSym, newFreq; - unsigned numPs = 0; - CPpmd_State *ps[PPMD7_MAX_ORDER]; - - if (p->OrderFall != 0) - ps[numPs++] = p->FoundState; - - while (c->Suffix) - { - CPpmd_Void_Ref successor; - CPpmd_State *s; - c = SUFFIX(c); - - - if (c->NumStats != 1) - { - Byte sym = p->FoundState->Symbol; - for (s = STATS(c); s->Symbol != sym; s++); - - } - else - { - s = ONE_STATE(c); - - } - successor = SUCCESSOR(s); - if (successor != upBranch) - { - // (c) is real record Context here, - c = CTX(successor); - if (numPs == 0) - { - // (c) is real record MAX Order Context here, - // So we don't need to create any new contexts. - return c; - } - break; - } - ps[numPs++] = s; - } - - // All created contexts will have single-symbol with new RAW-Successor - // All new RAW-Successors will point to next position in RAW text - // after FoundState->Successor - - newSym = *(const Byte *)Ppmd7_GetPtr(p, upBranch); - upBranch++; - - - if (c->NumStats == 1) - newFreq = ONE_STATE(c)->Freq; - else - { - UInt32 cf, s0; - CPpmd_State *s; - for (s = STATS(c); s->Symbol != newSym; s++); - cf = (UInt32)s->Freq - 1; - s0 = (UInt32)c->Union2.SummFreq - c->NumStats - cf; - /* - cf - is frequency of symbol that will be Successor in new context records. - s0 - is commulative frequency sum of another symbols from parent context. - max(newFreq)= (s->Freq + 1), when (s0 == 1) - we have requirement (Ppmd7Context_OneState()->Freq <= 128) in BinSumm[] - so (s->Freq < 128) - is requirement for multi-symbol contexts - */ - newFreq = (Byte)(1 + ((2 * cf <= s0) ? (5 * cf > s0) : (2 * cf + s0 - 1) / (2 * s0) + 1)); - } - - // Create new single-symbol contexts from low order to high order in loop - - do - { - CTX_PTR c1; - /* = AllocContext(p); */ - if (p->HiUnit != p->LoUnit) - c1 = (CTX_PTR)(void *)(p->HiUnit -= UNIT_SIZE); - else if (p->FreeList[0] != 0) - c1 = (CTX_PTR)RemoveNode(p, 0); - else - { - c1 = (CTX_PTR)AllocUnitsRare(p, 0); - if (!c1) - return NULL; - } - - c1->NumStats = 1; - ONE_STATE(c1)->Symbol = newSym; - ONE_STATE(c1)->Freq = newFreq; - SetSuccessor(ONE_STATE(c1), upBranch); - c1->Suffix = REF(c); - SetSuccessor(ps[--numPs], REF(c1)); - c = c1; - } - while (numPs != 0); - - return c; -} - - - -#define SwapStates(s) \ - { CPpmd_State tmp = s[0]; s[0] = s[-1]; s[-1] = tmp; } - - -void Ppmd7_UpdateModel(CPpmd7 *p); -MY_NO_INLINE -void Ppmd7_UpdateModel(CPpmd7 *p) -{ - CPpmd_Void_Ref maxSuccessor, minSuccessor; - CTX_PTR c, mc; - unsigned s0, ns; - - - - if (p->FoundState->Freq < MAX_FREQ / 4 && p->MinContext->Suffix != 0) - { - /* Update Freqs in Suffix Context */ - - c = SUFFIX(p->MinContext); - - if (c->NumStats == 1) - { - CPpmd_State *s = ONE_STATE(c); - if (s->Freq < 32) - s->Freq++; - } - else - { - CPpmd_State *s = STATS(c); - Byte sym = p->FoundState->Symbol; - - if (s->Symbol != sym) - { - do - { - // s++; if (s->Symbol == sym) break; - s++; - } - while (s->Symbol != sym); - - if (s[0].Freq >= s[-1].Freq) - { - SwapStates(s); - s--; - } - } - - if (s->Freq < MAX_FREQ - 9) - { - s->Freq = (Byte)(s->Freq + 2); - c->Union2.SummFreq = (UInt16)(c->Union2.SummFreq + 2); - } - } - } - - - if (p->OrderFall == 0) - { - /* MAX ORDER context */ - /* (FoundState->Successor) is RAW-Successor. */ - p->MaxContext = p->MinContext = CreateSuccessors(p); - if (!p->MinContext) - { - RestartModel(p); - return; - } - SetSuccessor(p->FoundState, REF(p->MinContext)); - return; - } - - - /* NON-MAX ORDER context */ - - { - Byte *text = p->Text; - *text++ = p->FoundState->Symbol; - p->Text = text; - if (text >= p->UnitsStart) - { - RestartModel(p); - return; - } - maxSuccessor = REF(text); - } - - minSuccessor = SUCCESSOR(p->FoundState); - - if (minSuccessor) - { - // there is Successor for FoundState in MinContext. - // So the next context will be one order higher than MinContext. - - if (minSuccessor <= maxSuccessor) - { - // minSuccessor is RAW-Successor. So we will create real contexts records: - CTX_PTR cs = CreateSuccessors(p); - if (!cs) - { - RestartModel(p); - return; - } - minSuccessor = REF(cs); - } - - // minSuccessor now is real Context pointer that points to existing (Order+1) context - - if (--p->OrderFall == 0) - { - /* - if we move to MaxOrder context, then minSuccessor will be common Succesor for both: - MinContext that is (MaxOrder - 1) - MaxContext that is (MaxOrder) - so we don't need new RAW-Successor, and we can use real minSuccessor - as succssors for both MinContext and MaxContext. - */ - maxSuccessor = minSuccessor; - - /* - if (MaxContext != MinContext) - { - there was order fall from MaxOrder and we don't need current symbol - to transfer some RAW-Succesors to real contexts. - So we roll back pointer in raw data for one position. - } - */ - p->Text -= (p->MaxContext != p->MinContext); - } - } - else - { - /* - FoundState has NULL-Successor here. - And only root 0-order context can contain NULL-Successors. - We change Successor in FoundState to RAW-Successor, - And next context will be same 0-order root Context. - */ - SetSuccessor(p->FoundState, maxSuccessor); - minSuccessor = REF(p->MinContext); - } - - mc = p->MinContext; - c = p->MaxContext; - - p->MaxContext = p->MinContext = CTX(minSuccessor); - - if (c == mc) - return; - - // s0 : is pure Escape Freq - s0 = mc->Union2.SummFreq - (ns = mc->NumStats) - ((unsigned)p->FoundState->Freq - 1); - - do - { - unsigned ns1; - UInt32 sum; - - if ((ns1 = c->NumStats) != 1) - { - if ((ns1 & 1) == 0) - { - /* Expand for one UNIT */ - unsigned oldNU = ns1 >> 1; - unsigned i = U2I(oldNU); - if (i != U2I((size_t)oldNU + 1)) - { - void *ptr = AllocUnits(p, i + 1); - void *oldPtr; - if (!ptr) - { - RestartModel(p); - return; - } - oldPtr = STATS(c); - MyMem12Cpy(ptr, oldPtr, oldNU); - InsertNode(p, oldPtr, i); - c->Union4.Stats = STATS_REF(ptr); - } - } - sum = c->Union2.SummFreq; - /* max increase of Escape_Freq is 3 here. - total increase of Union2.SummFreq for all symbols is less than 256 here */ - sum += (UInt32)(2 * ns1 < ns) + 2 * ((unsigned)(4 * ns1 <= ns) & (sum <= 8 * ns1)); - /* original PPMdH uses 16-bit variable for (sum) here. - But (sum < 0x9000). So we don't truncate (sum) to 16-bit */ - // sum = (UInt16)sum; - } - else - { - // instead of One-symbol context we create 2-symbol context - CPpmd_State *s = (CPpmd_State*)AllocUnits(p, 0); - if (!s) - { - RestartModel(p); - return; - } - { - unsigned freq = c->Union2.State2.Freq; - // s = *ONE_STATE(c); - s->Symbol = c->Union2.State2.Symbol; - s->Successor_0 = c->Union4.State4.Successor_0; - s->Successor_1 = c->Union4.State4.Successor_1; - // SetSuccessor(s, c->Union4.Stats); // call it only for debug purposes to check the order of - // (Successor_0 and Successor_1) in LE/BE. - c->Union4.Stats = REF(s); - if (freq < MAX_FREQ / 4 - 1) - freq <<= 1; - else - freq = MAX_FREQ - 4; - // (max(s->freq) == 120), when we convert from 1-symbol into 2-symbol context - s->Freq = (Byte)freq; - // max(InitEsc = PPMD7_kExpEscape[*]) is 25. So the max(escapeFreq) is 26 here - sum = freq + p->InitEsc + (ns > 3); - } - } - - { - CPpmd_State *s = STATS(c) + ns1; - UInt32 cf = 2 * (sum + 6) * (UInt32)p->FoundState->Freq; - UInt32 sf = (UInt32)s0 + sum; - s->Symbol = p->FoundState->Symbol; - c->NumStats = (UInt16)(ns1 + 1); - SetSuccessor(s, maxSuccessor); - - if (cf < 6 * sf) - { - cf = (UInt32)1 + (cf > sf) + (cf >= 4 * sf); - sum += 3; - /* It can add (0, 1, 2) to Escape_Freq */ - } - else - { - cf = (UInt32)4 + (cf >= 9 * sf) + (cf >= 12 * sf) + (cf >= 15 * sf); - sum += cf; - } - - c->Union2.SummFreq = (UInt16)sum; - s->Freq = (Byte)cf; - } - c = SUFFIX(c); - } - while (c != mc); -} - - - -MY_NO_INLINE -static void Rescale(CPpmd7 *p) -{ - unsigned i, adder, sumFreq, escFreq; - CPpmd_State *stats = STATS(p->MinContext); - CPpmd_State *s = p->FoundState; - - /* Sort the list by Freq */ - if (s != stats) - { - CPpmd_State tmp = *s; - do - s[0] = s[-1]; - while (--s != stats); - *s = tmp; - } - - sumFreq = s->Freq; - escFreq = p->MinContext->Union2.SummFreq - sumFreq; - - /* - if (p->OrderFall == 0), adder = 0 : it's allowed to remove symbol from MAX Order context - if (p->OrderFall != 0), adder = 1 : it's NOT allowed to remove symbol from NON-MAX Order context - */ - - adder = (p->OrderFall != 0); - - #ifdef PPMD7_ORDER_0_SUPPPORT - adder |= (p->MaxOrder == 0); // we don't remove symbols from order-0 context - #endif - - sumFreq = (sumFreq + 4 + adder) >> 1; - i = (unsigned)p->MinContext->NumStats - 1; - s->Freq = (Byte)sumFreq; - - do - { - unsigned freq = (++s)->Freq; - escFreq -= freq; - freq = (freq + adder) >> 1; - sumFreq += freq; - s->Freq = (Byte)freq; - if (freq > s[-1].Freq) - { - CPpmd_State tmp = *s; - CPpmd_State *s1 = s; - do - { - s1[0] = s1[-1]; - } - while (--s1 != stats && freq > s1[-1].Freq); - *s1 = tmp; - } - } - while (--i); - - if (s->Freq == 0) - { - /* Remove all items with Freq == 0 */ - CPpmd7_Context *mc; - unsigned numStats, numStatsNew, n0, n1; - - i = 0; do { i++; } while ((--s)->Freq == 0); - - /* We increase (escFreq) for the number of removed symbols. - So we will have (0.5) increase for Escape_Freq in avarage per - removed symbol after Escape_Freq halving */ - escFreq += i; - mc = p->MinContext; - numStats = mc->NumStats; - numStatsNew = numStats - i; - mc->NumStats = (UInt16)(numStatsNew); - n0 = (numStats + 1) >> 1; - - if (numStatsNew == 1) - { - /* Create Single-Symbol context */ - unsigned freq = stats->Freq; - - do - { - escFreq >>= 1; - freq = (freq + 1) >> 1; - } - while (escFreq > 1); - - s = ONE_STATE(mc); - *s = *stats; - s->Freq = (Byte)freq; // (freq <= 260 / 4) - p->FoundState = s; - InsertNode(p, stats, U2I(n0)); - return; - } - - n1 = (numStatsNew + 1) >> 1; - if (n0 != n1) - { - // p->MinContext->Union4.Stats = STATS_REF(ShrinkUnits(p, stats, n0, n1)); - unsigned i0 = U2I(n0); - unsigned i1 = U2I(n1); - if (i0 != i1) - { - if (p->FreeList[i1] != 0) - { - void *ptr = RemoveNode(p, i1); - p->MinContext->Union4.Stats = STATS_REF(ptr); - MyMem12Cpy(ptr, (const void *)stats, n1); - InsertNode(p, stats, i0); - } - else - SplitBlock(p, stats, i0, i1); - } - } - } - { - CPpmd7_Context *mc = p->MinContext; - mc->Union2.SummFreq = (UInt16)(sumFreq + escFreq - (escFreq >> 1)); - // Escape_Freq halving here - p->FoundState = STATS(mc); - } -} - - -CPpmd_See *Ppmd7_MakeEscFreq(CPpmd7 *p, unsigned numMasked, UInt32 *escFreq) -{ - CPpmd_See *see; - const CPpmd7_Context *mc = p->MinContext; - unsigned numStats = mc->NumStats; - if (numStats != 256) - { - unsigned nonMasked = numStats - numMasked; - see = p->See[(unsigned)p->NS2Indx[(size_t)nonMasked - 1]] - + (nonMasked < (unsigned)SUFFIX(mc)->NumStats - numStats) - + 2 * (unsigned)(mc->Union2.SummFreq < 11 * numStats) - + 4 * (unsigned)(numMasked > nonMasked) + - p->HiBitsFlag; - { - // if (see->Summ) field is larger than 16-bit, we need only low 16 bits of Summ - unsigned summ = (UInt16)see->Summ; // & 0xFFFF - unsigned r = (summ >> see->Shift); - see->Summ = (UInt16)(summ - r); - *escFreq = r + (r == 0); - } - } - else - { - see = &p->DummySee; - *escFreq = 1; - } - return see; -} - - -static void NextContext(CPpmd7 *p) -{ - CTX_PTR c = CTX(SUCCESSOR(p->FoundState)); - if (p->OrderFall == 0 && (const Byte *)c > p->Text) - p->MaxContext = p->MinContext = c; - else - Ppmd7_UpdateModel(p); -} - - -void Ppmd7_Update1(CPpmd7 *p) -{ - CPpmd_State *s = p->FoundState; - unsigned freq = s->Freq; - freq += 4; - p->MinContext->Union2.SummFreq = (UInt16)(p->MinContext->Union2.SummFreq + 4); - s->Freq = (Byte)freq; - if (freq > s[-1].Freq) - { - SwapStates(s); - p->FoundState = --s; - if (freq > MAX_FREQ) - Rescale(p); - } - NextContext(p); -} - - -void Ppmd7_Update1_0(CPpmd7 *p) -{ - CPpmd_State *s = p->FoundState; - CPpmd7_Context *mc = p->MinContext; - unsigned freq = s->Freq; - unsigned summFreq = mc->Union2.SummFreq; - p->PrevSuccess = (2 * freq > summFreq); - p->RunLength += (int)p->PrevSuccess; - mc->Union2.SummFreq = (UInt16)(summFreq + 4); - freq += 4; - s->Freq = (Byte)freq; - if (freq > MAX_FREQ) - Rescale(p); - NextContext(p); -} - - -/* -void Ppmd7_UpdateBin(CPpmd7 *p) -{ - unsigned freq = p->FoundState->Freq; - p->FoundState->Freq = (Byte)(freq + (freq < 128)); - p->PrevSuccess = 1; - p->RunLength++; - NextContext(p); -} -*/ - -void Ppmd7_Update2(CPpmd7 *p) -{ - CPpmd_State *s = p->FoundState; - unsigned freq = s->Freq; - freq += 4; - p->RunLength = p->InitRL; - p->MinContext->Union2.SummFreq = (UInt16)(p->MinContext->Union2.SummFreq + 4); - s->Freq = (Byte)freq; - if (freq > MAX_FREQ) - Rescale(p); - Ppmd7_UpdateModel(p); -} - - - -/* -PPMd Memory Map: -{ - [ 0 ] contains subset of original raw text, that is required to create context - records, Some symbols are not written, when max order context was reached - [ Text ] free area - [ UnitsStart ] CPpmd_State vectors and CPpmd7_Context records - [ LoUnit ] free area for CPpmd_State and CPpmd7_Context items -[ HiUnit ] CPpmd7_Context records - [ Size ] end of array -} - -These addresses don't cross at any time. -And the following condtions is true for addresses: - (0 <= Text < UnitsStart <= LoUnit <= HiUnit <= Size) - -Raw text is BYTE--aligned. -the data in block [ UnitsStart ... Size ] contains 12-bytes aligned UNITs. - -Last UNIT of array at offset (Size - 12) is root order-0 CPpmd7_Context record. -The code can free UNITs memory blocks that were allocated to store CPpmd_State vectors. -The code doesn't free UNITs allocated for CPpmd7_Context records. - -The code calls RestartModel(), when there is no free memory for allocation. -And RestartModel() changes the state to orignal start state, with full free block. - - -The code allocates UNITs with the following order: - -Allocation of 1 UNIT for Context record - - from free space (HiUnit) down to (LoUnit) - - from FreeList[0] - - AllocUnitsRare() - -AllocUnits() for CPpmd_State vectors: - - from FreeList[i] - - from free space (LoUnit) up to (HiUnit) - - AllocUnitsRare() - -AllocUnitsRare() - - if (GlueCount == 0) - { Glue lists, GlueCount = 255, allocate from FreeList[i]] } - - loop for all higher sized FreeList[...] lists - - from (UnitsStart - Text), GlueCount-- - - ERROR - - -Each Record with Context contains the CPpmd_State vector, where each -CPpmd_State contains the link to Successor. -There are 3 types of Successor: - 1) NULL-Successor - NULL pointer. NULL-Successor links can be stored - only in 0-order Root Context Record. - We use 0 value as NULL-Successor - 2) RAW-Successor - the link to position in raw text, - that "RAW-Successor" is being created after first - occurrence of new symbol for some existing context record. - (RAW-Successor > 0). - 3) RECORD-Successor - the link to CPpmd7_Context record of (Order+1), - that record is being created when we go via RAW-Successor again. - -For any successors at any time: the following condtions are true for Successor links: -(NULL-Successor < RAW-Successor < UnitsStart <= RECORD-Successor) - - ----------- Symbol Frequency, SummFreq and Range in Range_Coder ---------- - -CPpmd7_Context::SummFreq = Sum(Stats[].Freq) + Escape_Freq - -The PPMd code tries to fulfill the condition: - (SummFreq <= (256 * 128 = RC::kBot)) - -We have (Sum(Stats[].Freq) <= 256 * 124), because of (MAX_FREQ = 124) -So (4 = 128 - 124) is average reserve for Escape_Freq for each symbol. -If (CPpmd_State::Freq) is not aligned for 4, the reserve can be 5, 6 or 7. -SummFreq and Escape_Freq can be changed in Rescale() and *Update*() functions. -Rescale() can remove symbols only from max-order contexts. So Escape_Freq can increase after multiple calls of Rescale() for -max-order context. - -When the PPMd code still break (Total <= RC::Range) condition in range coder, -we have two ways to resolve that problem: - 1) we can report error, if we want to keep compatibility with original PPMd code that has no fix for such cases. - 2) we can reduce (Total) value to (RC::Range) by reducing (Escape_Freq) part of (Total) value. -*/ +/* Ppmd7.c -- PPMdH codec +2021-04-13 : Igor Pavlov : Public domain +This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */ + +#include "Precomp.h" + +#include + +#include "Ppmd7.h" + +/* define PPMD7_ORDER_0_SUPPPORT to suport order-0 mode, unsupported by orignal PPMd var.H. code */ +// #define PPMD7_ORDER_0_SUPPPORT + +MY_ALIGN(16) +static const Byte PPMD7_kExpEscape[16] = { 25, 14, 9, 7, 5, 5, 4, 4, 4, 3, 3, 3, 2, 2, 2, 2 }; +MY_ALIGN(16) +static const UInt16 kInitBinEsc[] = { 0x3CDD, 0x1F3F, 0x59BF, 0x48F3, 0x64A1, 0x5ABC, 0x6632, 0x6051}; + +#define MAX_FREQ 124 +#define UNIT_SIZE 12 + +#define U2B(nu) ((UInt32)(nu) * UNIT_SIZE) +#define U2I(nu) (p->Units2Indx[(size_t)(nu) - 1]) +#define I2U(indx) ((unsigned)p->Indx2Units[indx]) +#define I2U_UInt16(indx) ((UInt16)p->Indx2Units[indx]) + +#define REF(ptr) Ppmd_GetRef(p, ptr) + +#define STATS_REF(ptr) ((CPpmd_State_Ref)REF(ptr)) + +#define CTX(ref) ((CPpmd7_Context *)Ppmd7_GetContext(p, ref)) +#define STATS(ctx) Ppmd7_GetStats(p, ctx) +#define ONE_STATE(ctx) Ppmd7Context_OneState(ctx) +#define SUFFIX(ctx) CTX((ctx)->Suffix) + +typedef CPpmd7_Context * CTX_PTR; + +struct CPpmd7_Node_; + +typedef Ppmd_Ref_Type(struct CPpmd7_Node_) CPpmd7_Node_Ref; + +typedef struct CPpmd7_Node_ +{ + UInt16 Stamp; /* must be at offset 0 as CPpmd7_Context::NumStats. Stamp=0 means free */ + UInt16 NU; + CPpmd7_Node_Ref Next; /* must be at offset >= 4 */ + CPpmd7_Node_Ref Prev; +} CPpmd7_Node; + +#define NODE(r) Ppmd_GetPtr_Type(p, r, CPpmd7_Node) + +void Ppmd7_Construct(CPpmd7 *p) +{ + unsigned i, k, m; + + p->Base = NULL; + + for (i = 0, k = 0; i < PPMD_NUM_INDEXES; i++) + { + unsigned step = (i >= 12 ? 4 : (i >> 2) + 1); + do { p->Units2Indx[k++] = (Byte)i; } while (--step); + p->Indx2Units[i] = (Byte)k; + } + + p->NS2BSIndx[0] = (0 << 1); + p->NS2BSIndx[1] = (1 << 1); + memset(p->NS2BSIndx + 2, (2 << 1), 9); + memset(p->NS2BSIndx + 11, (3 << 1), 256 - 11); + + for (i = 0; i < 3; i++) + p->NS2Indx[i] = (Byte)i; + + for (m = i, k = 1; i < 256; i++) + { + p->NS2Indx[i] = (Byte)m; + if (--k == 0) + k = (++m) - 2; + } + + memcpy(p->ExpEscape, PPMD7_kExpEscape, 16); +} + + +void Ppmd7_Free(CPpmd7 *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->Base); + p->Size = 0; + p->Base = NULL; +} + + +BoolInt Ppmd7_Alloc(CPpmd7 *p, UInt32 size, ISzAllocPtr alloc) +{ + if (!p->Base || p->Size != size) + { + Ppmd7_Free(p, alloc); + p->AlignOffset = (4 - size) & 3; + if ((p->Base = (Byte *)ISzAlloc_Alloc(alloc, p->AlignOffset + size)) == NULL) + return False; + p->Size = size; + } + return True; +} + + + +// ---------- Internal Memory Allocator ---------- + +/* We can use CPpmd7_Node in list of free units (as in Ppmd8) + But we still need one additional list walk pass in GlueFreeBlocks(). + So we use simple CPpmd_Void_Ref instead of CPpmd7_Node in InsertNode() / RemoveNode() +*/ + +#define EMPTY_NODE 0 + + +static void InsertNode(CPpmd7 *p, void *node, unsigned indx) +{ + *((CPpmd_Void_Ref *)node) = p->FreeList[indx]; + // ((CPpmd7_Node *)node)->Next = (CPpmd7_Node_Ref)p->FreeList[indx]; + + p->FreeList[indx] = REF(node); + +} + + +static void *RemoveNode(CPpmd7 *p, unsigned indx) +{ + CPpmd_Void_Ref *node = (CPpmd_Void_Ref *)Ppmd7_GetPtr(p, p->FreeList[indx]); + p->FreeList[indx] = *node; + // CPpmd7_Node *node = NODE((CPpmd7_Node_Ref)p->FreeList[indx]); + // p->FreeList[indx] = node->Next; + return node; +} + + +static void SplitBlock(CPpmd7 *p, void *ptr, unsigned oldIndx, unsigned newIndx) +{ + unsigned i, nu = I2U(oldIndx) - I2U(newIndx); + ptr = (Byte *)ptr + U2B(I2U(newIndx)); + if (I2U(i = U2I(nu)) != nu) + { + unsigned k = I2U(--i); + InsertNode(p, ((Byte *)ptr) + U2B(k), nu - k - 1); + } + InsertNode(p, ptr, i); +} + + +/* we use CPpmd7_Node_Union union to solve XLC -O2 strict pointer aliasing problem */ + +typedef union _CPpmd7_Node_Union +{ + CPpmd7_Node Node; + CPpmd7_Node_Ref NextRef; +} CPpmd7_Node_Union; + +/* Original PPmdH (Ppmd7) code uses doubly linked list in GlueFreeBlocks() + we use single linked list similar to Ppmd8 code */ + + +static void GlueFreeBlocks(CPpmd7 *p) +{ + /* + we use first UInt16 field of 12-bytes UNITs as record type stamp + CPpmd_State { Byte Symbol; Byte Freq; : Freq != 0 + CPpmd7_Context { UInt16 NumStats; : NumStats != 0 + CPpmd7_Node { UInt16 Stamp : Stamp == 0 for free record + : Stamp == 1 for head record and guard + Last 12-bytes UNIT in array is always contains 12-bytes order-0 CPpmd7_Context record. + */ + CPpmd7_Node_Ref head, n = 0; + + p->GlueCount = 255; + + + /* we set guard NODE at LoUnit */ + if (p->LoUnit != p->HiUnit) + ((CPpmd7_Node *)(void *)p->LoUnit)->Stamp = 1; + + { + /* Create list of free blocks. + We still need one additional list walk pass before Glue. */ + unsigned i; + for (i = 0; i < PPMD_NUM_INDEXES; i++) + { + const UInt16 nu = I2U_UInt16(i); + CPpmd7_Node_Ref next = (CPpmd7_Node_Ref)p->FreeList[i]; + p->FreeList[i] = 0; + while (next != 0) + { + /* Don't change the order of the following commands: */ + CPpmd7_Node_Union *un = (CPpmd7_Node_Union *)NODE(next); + const CPpmd7_Node_Ref tmp = next; + next = un->NextRef; + un->Node.Stamp = EMPTY_NODE; + un->Node.NU = nu; + un->Node.Next = n; + n = tmp; + } + } + } + + head = n; + /* Glue and Fill must walk the list in same direction */ + { + /* Glue free blocks */ + CPpmd7_Node_Ref *prev = &head; + while (n) + { + CPpmd7_Node *node = NODE(n); + UInt32 nu = node->NU; + n = node->Next; + if (nu == 0) + { + *prev = n; + continue; + } + prev = &node->Next; + for (;;) + { + CPpmd7_Node *node2 = node + nu; + nu += node2->NU; + if (node2->Stamp != EMPTY_NODE || nu >= 0x10000) + break; + node->NU = (UInt16)nu; + node2->NU = 0; + } + } + } + + /* Fill lists of free blocks */ + for (n = head; n != 0;) + { + CPpmd7_Node *node = NODE(n); + UInt32 nu = node->NU; + unsigned i; + n = node->Next; + if (nu == 0) + continue; + for (; nu > 128; nu -= 128, node += 128) + InsertNode(p, node, PPMD_NUM_INDEXES - 1); + if (I2U(i = U2I(nu)) != nu) + { + unsigned k = I2U(--i); + InsertNode(p, node + k, (unsigned)nu - k - 1); + } + InsertNode(p, node, i); + } +} + + +MY_NO_INLINE +static void *AllocUnitsRare(CPpmd7 *p, unsigned indx) +{ + unsigned i; + + if (p->GlueCount == 0) + { + GlueFreeBlocks(p); + if (p->FreeList[indx] != 0) + return RemoveNode(p, indx); + } + + i = indx; + + do + { + if (++i == PPMD_NUM_INDEXES) + { + UInt32 numBytes = U2B(I2U(indx)); + Byte *us = p->UnitsStart; + p->GlueCount--; + return ((UInt32)(us - p->Text) > numBytes) ? (p->UnitsStart = us - numBytes) : NULL; + } + } + while (p->FreeList[i] == 0); + + { + void *block = RemoveNode(p, i); + SplitBlock(p, block, i, indx); + return block; + } +} + + +static void *AllocUnits(CPpmd7 *p, unsigned indx) +{ + if (p->FreeList[indx] != 0) + return RemoveNode(p, indx); + { + UInt32 numBytes = U2B(I2U(indx)); + Byte *lo = p->LoUnit; + if ((UInt32)(p->HiUnit - lo) >= numBytes) + { + p->LoUnit = lo + numBytes; + return lo; + } + } + return AllocUnitsRare(p, indx); +} + + +#define MyMem12Cpy(dest, src, num) \ + { UInt32 *d = (UInt32 *)dest; const UInt32 *z = (const UInt32 *)src; UInt32 n = num; \ + do { d[0] = z[0]; d[1] = z[1]; d[2] = z[2]; z += 3; d += 3; } while (--n); } + + +/* +static void *ShrinkUnits(CPpmd7 *p, void *oldPtr, unsigned oldNU, unsigned newNU) +{ + unsigned i0 = U2I(oldNU); + unsigned i1 = U2I(newNU); + if (i0 == i1) + return oldPtr; + if (p->FreeList[i1] != 0) + { + void *ptr = RemoveNode(p, i1); + MyMem12Cpy(ptr, oldPtr, newNU); + InsertNode(p, oldPtr, i0); + return ptr; + } + SplitBlock(p, oldPtr, i0, i1); + return oldPtr; +} +*/ + + +#define SUCCESSOR(p) Ppmd_GET_SUCCESSOR(p) +static void SetSuccessor(CPpmd_State *p, CPpmd_Void_Ref v) +{ + Ppmd_SET_SUCCESSOR(p, v); +} + + + +MY_NO_INLINE +static +void RestartModel(CPpmd7 *p) +{ + unsigned i, k; + + memset(p->FreeList, 0, sizeof(p->FreeList)); + + p->Text = p->Base + p->AlignOffset; + p->HiUnit = p->Text + p->Size; + p->LoUnit = p->UnitsStart = p->HiUnit - p->Size / 8 / UNIT_SIZE * 7 * UNIT_SIZE; + p->GlueCount = 0; + + p->OrderFall = p->MaxOrder; + p->RunLength = p->InitRL = -(Int32)((p->MaxOrder < 12) ? p->MaxOrder : 12) - 1; + p->PrevSuccess = 0; + + { + CPpmd7_Context *mc = (CTX_PTR)(void *)(p->HiUnit -= UNIT_SIZE); /* AllocContext(p); */ + CPpmd_State *s = (CPpmd_State *)p->LoUnit; /* AllocUnits(p, PPMD_NUM_INDEXES - 1); */ + + p->LoUnit += U2B(256 / 2); + p->MaxContext = p->MinContext = mc; + p->FoundState = s; + + mc->NumStats = 256; + mc->Union2.SummFreq = 256 + 1; + mc->Union4.Stats = REF(s); + mc->Suffix = 0; + + for (i = 0; i < 256; i++, s++) + { + s->Symbol = (Byte)i; + s->Freq = 1; + SetSuccessor(s, 0); + } + + #ifdef PPMD7_ORDER_0_SUPPPORT + if (p->MaxOrder == 0) + { + CPpmd_Void_Ref r = REF(mc); + s = p->FoundState; + for (i = 0; i < 256; i++, s++) + SetSuccessor(s, r); + return; + } + #endif + } + + for (i = 0; i < 128; i++) + + + + for (k = 0; k < 8; k++) + { + unsigned m; + UInt16 *dest = p->BinSumm[i] + k; + UInt16 val = (UInt16)(PPMD_BIN_SCALE - kInitBinEsc[k] / (i + 2)); + for (m = 0; m < 64; m += 8) + dest[m] = val; + } + + + for (i = 0; i < 25; i++) + { + + CPpmd_See *s = p->See[i]; + + + + unsigned summ = ((5 * i + 10) << (PPMD_PERIOD_BITS - 4)); + for (k = 0; k < 16; k++, s++) + { + s->Summ = (UInt16)summ; + s->Shift = (PPMD_PERIOD_BITS - 4); + s->Count = 4; + } + } + + p->DummySee.Summ = 0; /* unused */ + p->DummySee.Shift = PPMD_PERIOD_BITS; + p->DummySee.Count = 64; /* unused */ +} + + +void Ppmd7_Init(CPpmd7 *p, unsigned maxOrder) +{ + p->MaxOrder = maxOrder; + + RestartModel(p); +} + + + +/* + CreateSuccessors() + It's called when (FoundState->Successor) is RAW-Successor, + that is the link to position in Raw text. + So we create Context records and write the links to + FoundState->Successor and to identical RAW-Successors in suffix + contexts of MinContex. + + The function returns: + if (OrderFall == 0) then MinContext is already at MAX order, + { return pointer to new or existing context of same MAX order } + else + { return pointer to new real context that will be (Order+1) in comparison with MinContext + + also it can return pointer to real context of same order, +*/ + +MY_NO_INLINE +static CTX_PTR CreateSuccessors(CPpmd7 *p) +{ + CTX_PTR c = p->MinContext; + CPpmd_Byte_Ref upBranch = (CPpmd_Byte_Ref)SUCCESSOR(p->FoundState); + Byte newSym, newFreq; + unsigned numPs = 0; + CPpmd_State *ps[PPMD7_MAX_ORDER]; + + if (p->OrderFall != 0) + ps[numPs++] = p->FoundState; + + while (c->Suffix) + { + CPpmd_Void_Ref successor; + CPpmd_State *s; + c = SUFFIX(c); + + + if (c->NumStats != 1) + { + Byte sym = p->FoundState->Symbol; + for (s = STATS(c); s->Symbol != sym; s++); + + } + else + { + s = ONE_STATE(c); + + } + successor = SUCCESSOR(s); + if (successor != upBranch) + { + // (c) is real record Context here, + c = CTX(successor); + if (numPs == 0) + { + // (c) is real record MAX Order Context here, + // So we don't need to create any new contexts. + return c; + } + break; + } + ps[numPs++] = s; + } + + // All created contexts will have single-symbol with new RAW-Successor + // All new RAW-Successors will point to next position in RAW text + // after FoundState->Successor + + newSym = *(const Byte *)Ppmd7_GetPtr(p, upBranch); + upBranch++; + + + if (c->NumStats == 1) + newFreq = ONE_STATE(c)->Freq; + else + { + UInt32 cf, s0; + CPpmd_State *s; + for (s = STATS(c); s->Symbol != newSym; s++); + cf = (UInt32)s->Freq - 1; + s0 = (UInt32)c->Union2.SummFreq - c->NumStats - cf; + /* + cf - is frequency of symbol that will be Successor in new context records. + s0 - is commulative frequency sum of another symbols from parent context. + max(newFreq)= (s->Freq + 1), when (s0 == 1) + we have requirement (Ppmd7Context_OneState()->Freq <= 128) in BinSumm[] + so (s->Freq < 128) - is requirement for multi-symbol contexts + */ + newFreq = (Byte)(1 + ((2 * cf <= s0) ? (5 * cf > s0) : (2 * cf + s0 - 1) / (2 * s0) + 1)); + } + + // Create new single-symbol contexts from low order to high order in loop + + do + { + CTX_PTR c1; + /* = AllocContext(p); */ + if (p->HiUnit != p->LoUnit) + c1 = (CTX_PTR)(void *)(p->HiUnit -= UNIT_SIZE); + else if (p->FreeList[0] != 0) + c1 = (CTX_PTR)RemoveNode(p, 0); + else + { + c1 = (CTX_PTR)AllocUnitsRare(p, 0); + if (!c1) + return NULL; + } + + c1->NumStats = 1; + ONE_STATE(c1)->Symbol = newSym; + ONE_STATE(c1)->Freq = newFreq; + SetSuccessor(ONE_STATE(c1), upBranch); + c1->Suffix = REF(c); + SetSuccessor(ps[--numPs], REF(c1)); + c = c1; + } + while (numPs != 0); + + return c; +} + + + +#define SwapStates(s) \ + { CPpmd_State tmp = s[0]; s[0] = s[-1]; s[-1] = tmp; } + + +void Ppmd7_UpdateModel(CPpmd7 *p); +MY_NO_INLINE +void Ppmd7_UpdateModel(CPpmd7 *p) +{ + CPpmd_Void_Ref maxSuccessor, minSuccessor; + CTX_PTR c, mc; + unsigned s0, ns; + + + + if (p->FoundState->Freq < MAX_FREQ / 4 && p->MinContext->Suffix != 0) + { + /* Update Freqs in Suffix Context */ + + c = SUFFIX(p->MinContext); + + if (c->NumStats == 1) + { + CPpmd_State *s = ONE_STATE(c); + if (s->Freq < 32) + s->Freq++; + } + else + { + CPpmd_State *s = STATS(c); + Byte sym = p->FoundState->Symbol; + + if (s->Symbol != sym) + { + do + { + // s++; if (s->Symbol == sym) break; + s++; + } + while (s->Symbol != sym); + + if (s[0].Freq >= s[-1].Freq) + { + SwapStates(s); + s--; + } + } + + if (s->Freq < MAX_FREQ - 9) + { + s->Freq = (Byte)(s->Freq + 2); + c->Union2.SummFreq = (UInt16)(c->Union2.SummFreq + 2); + } + } + } + + + if (p->OrderFall == 0) + { + /* MAX ORDER context */ + /* (FoundState->Successor) is RAW-Successor. */ + p->MaxContext = p->MinContext = CreateSuccessors(p); + if (!p->MinContext) + { + RestartModel(p); + return; + } + SetSuccessor(p->FoundState, REF(p->MinContext)); + return; + } + + + /* NON-MAX ORDER context */ + + { + Byte *text = p->Text; + *text++ = p->FoundState->Symbol; + p->Text = text; + if (text >= p->UnitsStart) + { + RestartModel(p); + return; + } + maxSuccessor = REF(text); + } + + minSuccessor = SUCCESSOR(p->FoundState); + + if (minSuccessor) + { + // there is Successor for FoundState in MinContext. + // So the next context will be one order higher than MinContext. + + if (minSuccessor <= maxSuccessor) + { + // minSuccessor is RAW-Successor. So we will create real contexts records: + CTX_PTR cs = CreateSuccessors(p); + if (!cs) + { + RestartModel(p); + return; + } + minSuccessor = REF(cs); + } + + // minSuccessor now is real Context pointer that points to existing (Order+1) context + + if (--p->OrderFall == 0) + { + /* + if we move to MaxOrder context, then minSuccessor will be common Succesor for both: + MinContext that is (MaxOrder - 1) + MaxContext that is (MaxOrder) + so we don't need new RAW-Successor, and we can use real minSuccessor + as succssors for both MinContext and MaxContext. + */ + maxSuccessor = minSuccessor; + + /* + if (MaxContext != MinContext) + { + there was order fall from MaxOrder and we don't need current symbol + to transfer some RAW-Succesors to real contexts. + So we roll back pointer in raw data for one position. + } + */ + p->Text -= (p->MaxContext != p->MinContext); + } + } + else + { + /* + FoundState has NULL-Successor here. + And only root 0-order context can contain NULL-Successors. + We change Successor in FoundState to RAW-Successor, + And next context will be same 0-order root Context. + */ + SetSuccessor(p->FoundState, maxSuccessor); + minSuccessor = REF(p->MinContext); + } + + mc = p->MinContext; + c = p->MaxContext; + + p->MaxContext = p->MinContext = CTX(minSuccessor); + + if (c == mc) + return; + + // s0 : is pure Escape Freq + s0 = mc->Union2.SummFreq - (ns = mc->NumStats) - ((unsigned)p->FoundState->Freq - 1); + + do + { + unsigned ns1; + UInt32 sum; + + if ((ns1 = c->NumStats) != 1) + { + if ((ns1 & 1) == 0) + { + /* Expand for one UNIT */ + unsigned oldNU = ns1 >> 1; + unsigned i = U2I(oldNU); + if (i != U2I((size_t)oldNU + 1)) + { + void *ptr = AllocUnits(p, i + 1); + void *oldPtr; + if (!ptr) + { + RestartModel(p); + return; + } + oldPtr = STATS(c); + MyMem12Cpy(ptr, oldPtr, oldNU); + InsertNode(p, oldPtr, i); + c->Union4.Stats = STATS_REF(ptr); + } + } + sum = c->Union2.SummFreq; + /* max increase of Escape_Freq is 3 here. + total increase of Union2.SummFreq for all symbols is less than 256 here */ + sum += (UInt32)(2 * ns1 < ns) + 2 * ((unsigned)(4 * ns1 <= ns) & (sum <= 8 * ns1)); + /* original PPMdH uses 16-bit variable for (sum) here. + But (sum < 0x9000). So we don't truncate (sum) to 16-bit */ + // sum = (UInt16)sum; + } + else + { + // instead of One-symbol context we create 2-symbol context + CPpmd_State *s = (CPpmd_State*)AllocUnits(p, 0); + if (!s) + { + RestartModel(p); + return; + } + { + unsigned freq = c->Union2.State2.Freq; + // s = *ONE_STATE(c); + s->Symbol = c->Union2.State2.Symbol; + s->Successor_0 = c->Union4.State4.Successor_0; + s->Successor_1 = c->Union4.State4.Successor_1; + // SetSuccessor(s, c->Union4.Stats); // call it only for debug purposes to check the order of + // (Successor_0 and Successor_1) in LE/BE. + c->Union4.Stats = REF(s); + if (freq < MAX_FREQ / 4 - 1) + freq <<= 1; + else + freq = MAX_FREQ - 4; + // (max(s->freq) == 120), when we convert from 1-symbol into 2-symbol context + s->Freq = (Byte)freq; + // max(InitEsc = PPMD7_kExpEscape[*]) is 25. So the max(escapeFreq) is 26 here + sum = freq + p->InitEsc + (ns > 3); + } + } + + { + CPpmd_State *s = STATS(c) + ns1; + UInt32 cf = 2 * (sum + 6) * (UInt32)p->FoundState->Freq; + UInt32 sf = (UInt32)s0 + sum; + s->Symbol = p->FoundState->Symbol; + c->NumStats = (UInt16)(ns1 + 1); + SetSuccessor(s, maxSuccessor); + + if (cf < 6 * sf) + { + cf = (UInt32)1 + (cf > sf) + (cf >= 4 * sf); + sum += 3; + /* It can add (0, 1, 2) to Escape_Freq */ + } + else + { + cf = (UInt32)4 + (cf >= 9 * sf) + (cf >= 12 * sf) + (cf >= 15 * sf); + sum += cf; + } + + c->Union2.SummFreq = (UInt16)sum; + s->Freq = (Byte)cf; + } + c = SUFFIX(c); + } + while (c != mc); +} + + + +MY_NO_INLINE +static void Rescale(CPpmd7 *p) +{ + unsigned i, adder, sumFreq, escFreq; + CPpmd_State *stats = STATS(p->MinContext); + CPpmd_State *s = p->FoundState; + + /* Sort the list by Freq */ + if (s != stats) + { + CPpmd_State tmp = *s; + do + s[0] = s[-1]; + while (--s != stats); + *s = tmp; + } + + sumFreq = s->Freq; + escFreq = p->MinContext->Union2.SummFreq - sumFreq; + + /* + if (p->OrderFall == 0), adder = 0 : it's allowed to remove symbol from MAX Order context + if (p->OrderFall != 0), adder = 1 : it's NOT allowed to remove symbol from NON-MAX Order context + */ + + adder = (p->OrderFall != 0); + + #ifdef PPMD7_ORDER_0_SUPPPORT + adder |= (p->MaxOrder == 0); // we don't remove symbols from order-0 context + #endif + + sumFreq = (sumFreq + 4 + adder) >> 1; + i = (unsigned)p->MinContext->NumStats - 1; + s->Freq = (Byte)sumFreq; + + do + { + unsigned freq = (++s)->Freq; + escFreq -= freq; + freq = (freq + adder) >> 1; + sumFreq += freq; + s->Freq = (Byte)freq; + if (freq > s[-1].Freq) + { + CPpmd_State tmp = *s; + CPpmd_State *s1 = s; + do + { + s1[0] = s1[-1]; + } + while (--s1 != stats && freq > s1[-1].Freq); + *s1 = tmp; + } + } + while (--i); + + if (s->Freq == 0) + { + /* Remove all items with Freq == 0 */ + CPpmd7_Context *mc; + unsigned numStats, numStatsNew, n0, n1; + + i = 0; do { i++; } while ((--s)->Freq == 0); + + /* We increase (escFreq) for the number of removed symbols. + So we will have (0.5) increase for Escape_Freq in avarage per + removed symbol after Escape_Freq halving */ + escFreq += i; + mc = p->MinContext; + numStats = mc->NumStats; + numStatsNew = numStats - i; + mc->NumStats = (UInt16)(numStatsNew); + n0 = (numStats + 1) >> 1; + + if (numStatsNew == 1) + { + /* Create Single-Symbol context */ + unsigned freq = stats->Freq; + + do + { + escFreq >>= 1; + freq = (freq + 1) >> 1; + } + while (escFreq > 1); + + s = ONE_STATE(mc); + *s = *stats; + s->Freq = (Byte)freq; // (freq <= 260 / 4) + p->FoundState = s; + InsertNode(p, stats, U2I(n0)); + return; + } + + n1 = (numStatsNew + 1) >> 1; + if (n0 != n1) + { + // p->MinContext->Union4.Stats = STATS_REF(ShrinkUnits(p, stats, n0, n1)); + unsigned i0 = U2I(n0); + unsigned i1 = U2I(n1); + if (i0 != i1) + { + if (p->FreeList[i1] != 0) + { + void *ptr = RemoveNode(p, i1); + p->MinContext->Union4.Stats = STATS_REF(ptr); + MyMem12Cpy(ptr, (const void *)stats, n1); + InsertNode(p, stats, i0); + } + else + SplitBlock(p, stats, i0, i1); + } + } + } + { + CPpmd7_Context *mc = p->MinContext; + mc->Union2.SummFreq = (UInt16)(sumFreq + escFreq - (escFreq >> 1)); + // Escape_Freq halving here + p->FoundState = STATS(mc); + } +} + + +CPpmd_See *Ppmd7_MakeEscFreq(CPpmd7 *p, unsigned numMasked, UInt32 *escFreq) +{ + CPpmd_See *see; + const CPpmd7_Context *mc = p->MinContext; + unsigned numStats = mc->NumStats; + if (numStats != 256) + { + unsigned nonMasked = numStats - numMasked; + see = p->See[(unsigned)p->NS2Indx[(size_t)nonMasked - 1]] + + (nonMasked < (unsigned)SUFFIX(mc)->NumStats - numStats) + + 2 * (unsigned)(mc->Union2.SummFreq < 11 * numStats) + + 4 * (unsigned)(numMasked > nonMasked) + + p->HiBitsFlag; + { + // if (see->Summ) field is larger than 16-bit, we need only low 16 bits of Summ + unsigned summ = (UInt16)see->Summ; // & 0xFFFF + unsigned r = (summ >> see->Shift); + see->Summ = (UInt16)(summ - r); + *escFreq = r + (r == 0); + } + } + else + { + see = &p->DummySee; + *escFreq = 1; + } + return see; +} + + +static void NextContext(CPpmd7 *p) +{ + CTX_PTR c = CTX(SUCCESSOR(p->FoundState)); + if (p->OrderFall == 0 && (const Byte *)c > p->Text) + p->MaxContext = p->MinContext = c; + else + Ppmd7_UpdateModel(p); +} + + +void Ppmd7_Update1(CPpmd7 *p) +{ + CPpmd_State *s = p->FoundState; + unsigned freq = s->Freq; + freq += 4; + p->MinContext->Union2.SummFreq = (UInt16)(p->MinContext->Union2.SummFreq + 4); + s->Freq = (Byte)freq; + if (freq > s[-1].Freq) + { + SwapStates(s); + p->FoundState = --s; + if (freq > MAX_FREQ) + Rescale(p); + } + NextContext(p); +} + + +void Ppmd7_Update1_0(CPpmd7 *p) +{ + CPpmd_State *s = p->FoundState; + CPpmd7_Context *mc = p->MinContext; + unsigned freq = s->Freq; + unsigned summFreq = mc->Union2.SummFreq; + p->PrevSuccess = (2 * freq > summFreq); + p->RunLength += (int)p->PrevSuccess; + mc->Union2.SummFreq = (UInt16)(summFreq + 4); + freq += 4; + s->Freq = (Byte)freq; + if (freq > MAX_FREQ) + Rescale(p); + NextContext(p); +} + + +/* +void Ppmd7_UpdateBin(CPpmd7 *p) +{ + unsigned freq = p->FoundState->Freq; + p->FoundState->Freq = (Byte)(freq + (freq < 128)); + p->PrevSuccess = 1; + p->RunLength++; + NextContext(p); +} +*/ + +void Ppmd7_Update2(CPpmd7 *p) +{ + CPpmd_State *s = p->FoundState; + unsigned freq = s->Freq; + freq += 4; + p->RunLength = p->InitRL; + p->MinContext->Union2.SummFreq = (UInt16)(p->MinContext->Union2.SummFreq + 4); + s->Freq = (Byte)freq; + if (freq > MAX_FREQ) + Rescale(p); + Ppmd7_UpdateModel(p); +} + + + +/* +PPMd Memory Map: +{ + [ 0 ] contains subset of original raw text, that is required to create context + records, Some symbols are not written, when max order context was reached + [ Text ] free area + [ UnitsStart ] CPpmd_State vectors and CPpmd7_Context records + [ LoUnit ] free area for CPpmd_State and CPpmd7_Context items +[ HiUnit ] CPpmd7_Context records + [ Size ] end of array +} + +These addresses don't cross at any time. +And the following condtions is true for addresses: + (0 <= Text < UnitsStart <= LoUnit <= HiUnit <= Size) + +Raw text is BYTE--aligned. +the data in block [ UnitsStart ... Size ] contains 12-bytes aligned UNITs. + +Last UNIT of array at offset (Size - 12) is root order-0 CPpmd7_Context record. +The code can free UNITs memory blocks that were allocated to store CPpmd_State vectors. +The code doesn't free UNITs allocated for CPpmd7_Context records. + +The code calls RestartModel(), when there is no free memory for allocation. +And RestartModel() changes the state to orignal start state, with full free block. + + +The code allocates UNITs with the following order: + +Allocation of 1 UNIT for Context record + - from free space (HiUnit) down to (LoUnit) + - from FreeList[0] + - AllocUnitsRare() + +AllocUnits() for CPpmd_State vectors: + - from FreeList[i] + - from free space (LoUnit) up to (HiUnit) + - AllocUnitsRare() + +AllocUnitsRare() + - if (GlueCount == 0) + { Glue lists, GlueCount = 255, allocate from FreeList[i]] } + - loop for all higher sized FreeList[...] lists + - from (UnitsStart - Text), GlueCount-- + - ERROR + + +Each Record with Context contains the CPpmd_State vector, where each +CPpmd_State contains the link to Successor. +There are 3 types of Successor: + 1) NULL-Successor - NULL pointer. NULL-Successor links can be stored + only in 0-order Root Context Record. + We use 0 value as NULL-Successor + 2) RAW-Successor - the link to position in raw text, + that "RAW-Successor" is being created after first + occurrence of new symbol for some existing context record. + (RAW-Successor > 0). + 3) RECORD-Successor - the link to CPpmd7_Context record of (Order+1), + that record is being created when we go via RAW-Successor again. + +For any successors at any time: the following condtions are true for Successor links: +(NULL-Successor < RAW-Successor < UnitsStart <= RECORD-Successor) + + +---------- Symbol Frequency, SummFreq and Range in Range_Coder ---------- + +CPpmd7_Context::SummFreq = Sum(Stats[].Freq) + Escape_Freq + +The PPMd code tries to fulfill the condition: + (SummFreq <= (256 * 128 = RC::kBot)) + +We have (Sum(Stats[].Freq) <= 256 * 124), because of (MAX_FREQ = 124) +So (4 = 128 - 124) is average reserve for Escape_Freq for each symbol. +If (CPpmd_State::Freq) is not aligned for 4, the reserve can be 5, 6 or 7. +SummFreq and Escape_Freq can be changed in Rescale() and *Update*() functions. +Rescale() can remove symbols only from max-order contexts. So Escape_Freq can increase after multiple calls of Rescale() for +max-order context. + +When the PPMd code still break (Total <= RC::Range) condition in range coder, +we have two ways to resolve that problem: + 1) we can report error, if we want to keep compatibility with original PPMd code that has no fix for such cases. + 2) we can reduce (Total) value to (RC::Range) by reducing (Escape_Freq) part of (Total) value. +*/ diff --git a/deps/LZMA-SDK/C/Ppmd7.h b/deps/LZMA-SDK/C/Ppmd7.h index 297e35fe9..d31809aeb 100644 --- a/deps/LZMA-SDK/C/Ppmd7.h +++ b/deps/LZMA-SDK/C/Ppmd7.h @@ -1,181 +1,181 @@ -/* Ppmd7.h -- Ppmd7 (PPMdH) compression codec -2021-04-13 : Igor Pavlov : Public domain -This code is based on: - PPMd var.H (2001): Dmitry Shkarin : Public domain */ - - -#ifndef __PPMD7_H -#define __PPMD7_H - -#include "Ppmd.h" - -EXTERN_C_BEGIN - -#define PPMD7_MIN_ORDER 2 -#define PPMD7_MAX_ORDER 64 - -#define PPMD7_MIN_MEM_SIZE (1 << 11) -#define PPMD7_MAX_MEM_SIZE (0xFFFFFFFF - 12 * 3) - -struct CPpmd7_Context_; - -typedef Ppmd_Ref_Type(struct CPpmd7_Context_) CPpmd7_Context_Ref; - -// MY_CPU_pragma_pack_push_1 - -typedef struct CPpmd7_Context_ -{ - UInt16 NumStats; - - - union - { - UInt16 SummFreq; - CPpmd_State2 State2; - } Union2; - - union - { - CPpmd_State_Ref Stats; - CPpmd_State4 State4; - } Union4; - - CPpmd7_Context_Ref Suffix; -} CPpmd7_Context; - -// MY_CPU_pragma_pop - -#define Ppmd7Context_OneState(p) ((CPpmd_State *)&(p)->Union2) - - - - -typedef struct -{ - UInt32 Range; - UInt32 Code; - UInt32 Low; - IByteIn *Stream; -} CPpmd7_RangeDec; - - -typedef struct -{ - UInt32 Range; - Byte Cache; - // Byte _dummy_[3]; - UInt64 Low; - UInt64 CacheSize; - IByteOut *Stream; -} CPpmd7z_RangeEnc; - - -typedef struct -{ - CPpmd7_Context *MinContext, *MaxContext; - CPpmd_State *FoundState; - unsigned OrderFall, InitEsc, PrevSuccess, MaxOrder, HiBitsFlag; - Int32 RunLength, InitRL; /* must be 32-bit at least */ - - UInt32 Size; - UInt32 GlueCount; - UInt32 AlignOffset; - Byte *Base, *LoUnit, *HiUnit, *Text, *UnitsStart; - - - - - union - { - CPpmd7_RangeDec dec; - CPpmd7z_RangeEnc enc; - } rc; - - Byte Indx2Units[PPMD_NUM_INDEXES + 2]; // +2 for alignment - Byte Units2Indx[128]; - CPpmd_Void_Ref FreeList[PPMD_NUM_INDEXES]; - - Byte NS2BSIndx[256], NS2Indx[256]; - Byte ExpEscape[16]; - CPpmd_See DummySee, See[25][16]; - UInt16 BinSumm[128][64]; - // int LastSymbol; -} CPpmd7; - - -void Ppmd7_Construct(CPpmd7 *p); -BoolInt Ppmd7_Alloc(CPpmd7 *p, UInt32 size, ISzAllocPtr alloc); -void Ppmd7_Free(CPpmd7 *p, ISzAllocPtr alloc); -void Ppmd7_Init(CPpmd7 *p, unsigned maxOrder); -#define Ppmd7_WasAllocated(p) ((p)->Base != NULL) - - -/* ---------- Internal Functions ---------- */ - -#define Ppmd7_GetPtr(p, ptr) Ppmd_GetPtr(p, ptr) -#define Ppmd7_GetContext(p, ptr) Ppmd_GetPtr_Type(p, ptr, CPpmd7_Context) -#define Ppmd7_GetStats(p, ctx) Ppmd_GetPtr_Type(p, (ctx)->Union4.Stats, CPpmd_State) - -void Ppmd7_Update1(CPpmd7 *p); -void Ppmd7_Update1_0(CPpmd7 *p); -void Ppmd7_Update2(CPpmd7 *p); - -#define PPMD7_HiBitsFlag_3(sym) ((((unsigned)sym + 0xC0) >> (8 - 3)) & (1 << 3)) -#define PPMD7_HiBitsFlag_4(sym) ((((unsigned)sym + 0xC0) >> (8 - 4)) & (1 << 4)) -// #define PPMD7_HiBitsFlag_3(sym) ((sym) < 0x40 ? 0 : (1 << 3)) -// #define PPMD7_HiBitsFlag_4(sym) ((sym) < 0x40 ? 0 : (1 << 4)) - -#define Ppmd7_GetBinSumm(p) \ - &p->BinSumm[(size_t)(unsigned)Ppmd7Context_OneState(p->MinContext)->Freq - 1] \ - [ p->PrevSuccess + ((p->RunLength >> 26) & 0x20) \ - + p->NS2BSIndx[(size_t)Ppmd7_GetContext(p, p->MinContext->Suffix)->NumStats - 1] \ - + PPMD7_HiBitsFlag_4(Ppmd7Context_OneState(p->MinContext)->Symbol) \ - + (p->HiBitsFlag = PPMD7_HiBitsFlag_3(p->FoundState->Symbol)) ] - -CPpmd_See *Ppmd7_MakeEscFreq(CPpmd7 *p, unsigned numMasked, UInt32 *scale); - - -/* -We support two versions of Ppmd7 (PPMdH) methods that use same CPpmd7 structure: - 1) Ppmd7a_*: original PPMdH - 2) Ppmd7z_*: modified PPMdH with 7z Range Coder -Ppmd7_*: the structures and functions that are common for both versions of PPMd7 (PPMdH) -*/ - -/* ---------- Decode ---------- */ - -#define PPMD7_SYM_END (-1) -#define PPMD7_SYM_ERROR (-2) - -/* -You must set (CPpmd7::rc.dec.Stream) before Ppmd7*_RangeDec_Init() - -Ppmd7*_DecodeSymbol() -out: - >= 0 : decoded byte - -1 : PPMD7_SYM_END : End of payload marker - -2 : PPMD7_SYM_ERROR : Data error -*/ - -/* Ppmd7a_* : original PPMdH */ -BoolInt Ppmd7a_RangeDec_Init(CPpmd7_RangeDec *p); -#define Ppmd7a_RangeDec_IsFinishedOK(p) ((p)->Code == 0) -int Ppmd7a_DecodeSymbol(CPpmd7 *p); - -/* Ppmd7z_* : modified PPMdH with 7z Range Coder */ -BoolInt Ppmd7z_RangeDec_Init(CPpmd7_RangeDec *p); -#define Ppmd7z_RangeDec_IsFinishedOK(p) ((p)->Code == 0) -int Ppmd7z_DecodeSymbol(CPpmd7 *p); -// Byte *Ppmd7z_DecodeSymbols(CPpmd7 *p, Byte *buf, const Byte *lim); - - -/* ---------- Encode ---------- */ - -void Ppmd7z_Init_RangeEnc(CPpmd7 *p); -void Ppmd7z_Flush_RangeEnc(CPpmd7 *p); -// void Ppmd7z_EncodeSymbol(CPpmd7 *p, int symbol); -void Ppmd7z_EncodeSymbols(CPpmd7 *p, const Byte *buf, const Byte *lim); - -EXTERN_C_END - -#endif +/* Ppmd7.h -- Ppmd7 (PPMdH) compression codec +2021-04-13 : Igor Pavlov : Public domain +This code is based on: + PPMd var.H (2001): Dmitry Shkarin : Public domain */ + + +#ifndef __PPMD7_H +#define __PPMD7_H + +#include "Ppmd.h" + +EXTERN_C_BEGIN + +#define PPMD7_MIN_ORDER 2 +#define PPMD7_MAX_ORDER 64 + +#define PPMD7_MIN_MEM_SIZE (1 << 11) +#define PPMD7_MAX_MEM_SIZE (0xFFFFFFFF - 12 * 3) + +struct CPpmd7_Context_; + +typedef Ppmd_Ref_Type(struct CPpmd7_Context_) CPpmd7_Context_Ref; + +// MY_CPU_pragma_pack_push_1 + +typedef struct CPpmd7_Context_ +{ + UInt16 NumStats; + + + union + { + UInt16 SummFreq; + CPpmd_State2 State2; + } Union2; + + union + { + CPpmd_State_Ref Stats; + CPpmd_State4 State4; + } Union4; + + CPpmd7_Context_Ref Suffix; +} CPpmd7_Context; + +// MY_CPU_pragma_pop + +#define Ppmd7Context_OneState(p) ((CPpmd_State *)&(p)->Union2) + + + + +typedef struct +{ + UInt32 Range; + UInt32 Code; + UInt32 Low; + IByteIn *Stream; +} CPpmd7_RangeDec; + + +typedef struct +{ + UInt32 Range; + Byte Cache; + // Byte _dummy_[3]; + UInt64 Low; + UInt64 CacheSize; + IByteOut *Stream; +} CPpmd7z_RangeEnc; + + +typedef struct +{ + CPpmd7_Context *MinContext, *MaxContext; + CPpmd_State *FoundState; + unsigned OrderFall, InitEsc, PrevSuccess, MaxOrder, HiBitsFlag; + Int32 RunLength, InitRL; /* must be 32-bit at least */ + + UInt32 Size; + UInt32 GlueCount; + UInt32 AlignOffset; + Byte *Base, *LoUnit, *HiUnit, *Text, *UnitsStart; + + + + + union + { + CPpmd7_RangeDec dec; + CPpmd7z_RangeEnc enc; + } rc; + + Byte Indx2Units[PPMD_NUM_INDEXES + 2]; // +2 for alignment + Byte Units2Indx[128]; + CPpmd_Void_Ref FreeList[PPMD_NUM_INDEXES]; + + Byte NS2BSIndx[256], NS2Indx[256]; + Byte ExpEscape[16]; + CPpmd_See DummySee, See[25][16]; + UInt16 BinSumm[128][64]; + // int LastSymbol; +} CPpmd7; + + +void Ppmd7_Construct(CPpmd7 *p); +BoolInt Ppmd7_Alloc(CPpmd7 *p, UInt32 size, ISzAllocPtr alloc); +void Ppmd7_Free(CPpmd7 *p, ISzAllocPtr alloc); +void Ppmd7_Init(CPpmd7 *p, unsigned maxOrder); +#define Ppmd7_WasAllocated(p) ((p)->Base != NULL) + + +/* ---------- Internal Functions ---------- */ + +#define Ppmd7_GetPtr(p, ptr) Ppmd_GetPtr(p, ptr) +#define Ppmd7_GetContext(p, ptr) Ppmd_GetPtr_Type(p, ptr, CPpmd7_Context) +#define Ppmd7_GetStats(p, ctx) Ppmd_GetPtr_Type(p, (ctx)->Union4.Stats, CPpmd_State) + +void Ppmd7_Update1(CPpmd7 *p); +void Ppmd7_Update1_0(CPpmd7 *p); +void Ppmd7_Update2(CPpmd7 *p); + +#define PPMD7_HiBitsFlag_3(sym) ((((unsigned)sym + 0xC0) >> (8 - 3)) & (1 << 3)) +#define PPMD7_HiBitsFlag_4(sym) ((((unsigned)sym + 0xC0) >> (8 - 4)) & (1 << 4)) +// #define PPMD7_HiBitsFlag_3(sym) ((sym) < 0x40 ? 0 : (1 << 3)) +// #define PPMD7_HiBitsFlag_4(sym) ((sym) < 0x40 ? 0 : (1 << 4)) + +#define Ppmd7_GetBinSumm(p) \ + &p->BinSumm[(size_t)(unsigned)Ppmd7Context_OneState(p->MinContext)->Freq - 1] \ + [ p->PrevSuccess + ((p->RunLength >> 26) & 0x20) \ + + p->NS2BSIndx[(size_t)Ppmd7_GetContext(p, p->MinContext->Suffix)->NumStats - 1] \ + + PPMD7_HiBitsFlag_4(Ppmd7Context_OneState(p->MinContext)->Symbol) \ + + (p->HiBitsFlag = PPMD7_HiBitsFlag_3(p->FoundState->Symbol)) ] + +CPpmd_See *Ppmd7_MakeEscFreq(CPpmd7 *p, unsigned numMasked, UInt32 *scale); + + +/* +We support two versions of Ppmd7 (PPMdH) methods that use same CPpmd7 structure: + 1) Ppmd7a_*: original PPMdH + 2) Ppmd7z_*: modified PPMdH with 7z Range Coder +Ppmd7_*: the structures and functions that are common for both versions of PPMd7 (PPMdH) +*/ + +/* ---------- Decode ---------- */ + +#define PPMD7_SYM_END (-1) +#define PPMD7_SYM_ERROR (-2) + +/* +You must set (CPpmd7::rc.dec.Stream) before Ppmd7*_RangeDec_Init() + +Ppmd7*_DecodeSymbol() +out: + >= 0 : decoded byte + -1 : PPMD7_SYM_END : End of payload marker + -2 : PPMD7_SYM_ERROR : Data error +*/ + +/* Ppmd7a_* : original PPMdH */ +BoolInt Ppmd7a_RangeDec_Init(CPpmd7_RangeDec *p); +#define Ppmd7a_RangeDec_IsFinishedOK(p) ((p)->Code == 0) +int Ppmd7a_DecodeSymbol(CPpmd7 *p); + +/* Ppmd7z_* : modified PPMdH with 7z Range Coder */ +BoolInt Ppmd7z_RangeDec_Init(CPpmd7_RangeDec *p); +#define Ppmd7z_RangeDec_IsFinishedOK(p) ((p)->Code == 0) +int Ppmd7z_DecodeSymbol(CPpmd7 *p); +// Byte *Ppmd7z_DecodeSymbols(CPpmd7 *p, Byte *buf, const Byte *lim); + + +/* ---------- Encode ---------- */ + +void Ppmd7z_Init_RangeEnc(CPpmd7 *p); +void Ppmd7z_Flush_RangeEnc(CPpmd7 *p); +// void Ppmd7z_EncodeSymbol(CPpmd7 *p, int symbol); +void Ppmd7z_EncodeSymbols(CPpmd7 *p, const Byte *buf, const Byte *lim); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/Ppmd7Dec.c b/deps/LZMA-SDK/C/Ppmd7Dec.c index a18f0b873..55d74ff9d 100644 --- a/deps/LZMA-SDK/C/Ppmd7Dec.c +++ b/deps/LZMA-SDK/C/Ppmd7Dec.c @@ -1,297 +1,297 @@ -/* Ppmd7Dec.c -- Ppmd7z (PPMdH with 7z Range Coder) Decoder -2021-04-13 : Igor Pavlov : Public domain -This code is based on: - PPMd var.H (2001): Dmitry Shkarin : Public domain */ - - -#include "Precomp.h" - -#include "Ppmd7.h" - -#define kTopValue (1 << 24) - - -#define READ_BYTE(p) IByteIn_Read((p)->Stream) - -BoolInt Ppmd7z_RangeDec_Init(CPpmd7_RangeDec *p) -{ - unsigned i; - p->Code = 0; - p->Range = 0xFFFFFFFF; - if (READ_BYTE(p) != 0) - return False; - for (i = 0; i < 4; i++) - p->Code = (p->Code << 8) | READ_BYTE(p); - return (p->Code < 0xFFFFFFFF); -} - -#define RC_NORM_BASE(p) if ((p)->Range < kTopValue) \ - { (p)->Code = ((p)->Code << 8) | READ_BYTE(p); (p)->Range <<= 8; - -#define RC_NORM_1(p) RC_NORM_BASE(p) } -#define RC_NORM(p) RC_NORM_BASE(p) RC_NORM_BASE(p) }} - -// we must use only one type of Normalization from two: LOCAL or REMOTE -#define RC_NORM_LOCAL(p) // RC_NORM(p) -#define RC_NORM_REMOTE(p) RC_NORM(p) - -#define R (&p->rc.dec) - -MY_FORCE_INLINE -// MY_NO_INLINE -static void RangeDec_Decode(CPpmd7 *p, UInt32 start, UInt32 size) -{ - - - R->Code -= start * R->Range; - R->Range *= size; - RC_NORM_LOCAL(R) -} - -#define RC_Decode(start, size) RangeDec_Decode(p, start, size); -#define RC_DecodeFinal(start, size) RC_Decode(start, size) RC_NORM_REMOTE(R) -#define RC_GetThreshold(total) (R->Code / (R->Range /= (total))) - - -#define CTX(ref) ((CPpmd7_Context *)Ppmd7_GetContext(p, ref)) -typedef CPpmd7_Context * CTX_PTR; -#define SUCCESSOR(p) Ppmd_GET_SUCCESSOR(p) -void Ppmd7_UpdateModel(CPpmd7 *p); - -#define MASK(sym) ((unsigned char *)charMask)[sym] -// MY_FORCE_INLINE -// static -int Ppmd7z_DecodeSymbol(CPpmd7 *p) -{ - size_t charMask[256 / sizeof(size_t)]; - - if (p->MinContext->NumStats != 1) - { - CPpmd_State *s = Ppmd7_GetStats(p, p->MinContext); - unsigned i; - UInt32 count, hiCnt; - UInt32 summFreq = p->MinContext->Union2.SummFreq; - - - - - count = RC_GetThreshold(summFreq); - hiCnt = count; - - if ((Int32)(count -= s->Freq) < 0) - { - Byte sym; - RC_DecodeFinal(0, s->Freq); - p->FoundState = s; - sym = s->Symbol; - Ppmd7_Update1_0(p); - return sym; - } - - p->PrevSuccess = 0; - i = (unsigned)p->MinContext->NumStats - 1; - - do - { - if ((Int32)(count -= (++s)->Freq) < 0) - { - Byte sym; - RC_DecodeFinal((hiCnt - count) - s->Freq, s->Freq); - p->FoundState = s; - sym = s->Symbol; - Ppmd7_Update1(p); - return sym; - } - } - while (--i); - - if (hiCnt >= summFreq) - return PPMD7_SYM_ERROR; - - hiCnt -= count; - RC_Decode(hiCnt, summFreq - hiCnt); - - p->HiBitsFlag = PPMD7_HiBitsFlag_3(p->FoundState->Symbol); - PPMD_SetAllBitsIn256Bytes(charMask); - // i = p->MinContext->NumStats - 1; - // do { MASK((--s)->Symbol) = 0; } while (--i); - { - CPpmd_State *s2 = Ppmd7_GetStats(p, p->MinContext); - MASK(s->Symbol) = 0; - do - { - unsigned sym0 = s2[0].Symbol; - unsigned sym1 = s2[1].Symbol; - s2 += 2; - MASK(sym0) = 0; - MASK(sym1) = 0; - } - while (s2 < s); - } - } - else - { - CPpmd_State *s = Ppmd7Context_OneState(p->MinContext); - UInt16 *prob = Ppmd7_GetBinSumm(p); - UInt32 pr = *prob; - UInt32 size0 = (R->Range >> 14) * pr; - pr = PPMD_UPDATE_PROB_1(pr); - - if (R->Code < size0) - { - Byte sym; - *prob = (UInt16)(pr + (1 << PPMD_INT_BITS)); - - // RangeDec_DecodeBit0(size0); - R->Range = size0; - RC_NORM_1(R) - /* we can use single byte normalization here because of - (min(BinSumm[][]) = 95) > (1 << (14 - 8)) */ - - // sym = (p->FoundState = Ppmd7Context_OneState(p->MinContext))->Symbol; - // Ppmd7_UpdateBin(p); - { - unsigned freq = s->Freq; - CTX_PTR c = CTX(SUCCESSOR(s)); - sym = s->Symbol; - p->FoundState = s; - p->PrevSuccess = 1; - p->RunLength++; - s->Freq = (Byte)(freq + (freq < 128)); - // NextContext(p); - if (p->OrderFall == 0 && (const Byte *)c > p->Text) - p->MaxContext = p->MinContext = c; - else - Ppmd7_UpdateModel(p); - } - return sym; - } - - *prob = (UInt16)pr; - p->InitEsc = p->ExpEscape[pr >> 10]; - - // RangeDec_DecodeBit1(size0); - - R->Code -= size0; - R->Range -= size0; - RC_NORM_LOCAL(R) - - PPMD_SetAllBitsIn256Bytes(charMask); - MASK(Ppmd7Context_OneState(p->MinContext)->Symbol) = 0; - p->PrevSuccess = 0; - } - - for (;;) - { - CPpmd_State *s, *s2; - UInt32 freqSum, count, hiCnt; - - CPpmd_See *see; - CPpmd7_Context *mc; - unsigned numMasked; - RC_NORM_REMOTE(R) - mc = p->MinContext; - numMasked = mc->NumStats; - - do - { - p->OrderFall++; - if (!mc->Suffix) - return PPMD7_SYM_END; - mc = Ppmd7_GetContext(p, mc->Suffix); - } - while (mc->NumStats == numMasked); - - s = Ppmd7_GetStats(p, mc); - - { - unsigned num = mc->NumStats; - unsigned num2 = num / 2; - - num &= 1; - hiCnt = (s->Freq & (unsigned)(MASK(s->Symbol))) & (0 - (UInt32)num); - s += num; - p->MinContext = mc; - - do - { - unsigned sym0 = s[0].Symbol; - unsigned sym1 = s[1].Symbol; - s += 2; - hiCnt += (s[-2].Freq & (unsigned)(MASK(sym0))); - hiCnt += (s[-1].Freq & (unsigned)(MASK(sym1))); - } - while (--num2); - } - - see = Ppmd7_MakeEscFreq(p, numMasked, &freqSum); - freqSum += hiCnt; - - - - - count = RC_GetThreshold(freqSum); - - if (count < hiCnt) - { - Byte sym; - - s = Ppmd7_GetStats(p, p->MinContext); - hiCnt = count; - // count -= s->Freq & (unsigned)(MASK(s->Symbol)); - // if ((Int32)count >= 0) - { - for (;;) - { - count -= s->Freq & (unsigned)(MASK((s)->Symbol)); s++; if ((Int32)count < 0) break; - // count -= s->Freq & (unsigned)(MASK((s)->Symbol)); s++; if ((Int32)count < 0) break; - }; - } - s--; - RC_DecodeFinal((hiCnt - count) - s->Freq, s->Freq); - - // new (see->Summ) value can overflow over 16-bits in some rare cases - Ppmd_See_Update(see); - p->FoundState = s; - sym = s->Symbol; - Ppmd7_Update2(p); - return sym; - } - - if (count >= freqSum) - return PPMD7_SYM_ERROR; - - RC_Decode(hiCnt, freqSum - hiCnt); - - // We increase (see->Summ) for sum of Freqs of all non_Masked symbols. - // new (see->Summ) value can overflow over 16-bits in some rare cases - see->Summ = (UInt16)(see->Summ + freqSum); - - s = Ppmd7_GetStats(p, p->MinContext); - s2 = s + p->MinContext->NumStats; - do - { - MASK(s->Symbol) = 0; - s++; - } - while (s != s2); - } -} - -/* -Byte *Ppmd7z_DecodeSymbols(CPpmd7 *p, Byte *buf, const Byte *lim) -{ - int sym = 0; - if (buf != lim) - do - { - sym = Ppmd7z_DecodeSymbol(p); - if (sym < 0) - break; - *buf = (Byte)sym; - } - while (++buf < lim); - p->LastSymbol = sym; - return buf; -} -*/ +/* Ppmd7Dec.c -- Ppmd7z (PPMdH with 7z Range Coder) Decoder +2021-04-13 : Igor Pavlov : Public domain +This code is based on: + PPMd var.H (2001): Dmitry Shkarin : Public domain */ + + +#include "Precomp.h" + +#include "Ppmd7.h" + +#define kTopValue (1 << 24) + + +#define READ_BYTE(p) IByteIn_Read((p)->Stream) + +BoolInt Ppmd7z_RangeDec_Init(CPpmd7_RangeDec *p) +{ + unsigned i; + p->Code = 0; + p->Range = 0xFFFFFFFF; + if (READ_BYTE(p) != 0) + return False; + for (i = 0; i < 4; i++) + p->Code = (p->Code << 8) | READ_BYTE(p); + return (p->Code < 0xFFFFFFFF); +} + +#define RC_NORM_BASE(p) if ((p)->Range < kTopValue) \ + { (p)->Code = ((p)->Code << 8) | READ_BYTE(p); (p)->Range <<= 8; + +#define RC_NORM_1(p) RC_NORM_BASE(p) } +#define RC_NORM(p) RC_NORM_BASE(p) RC_NORM_BASE(p) }} + +// we must use only one type of Normalization from two: LOCAL or REMOTE +#define RC_NORM_LOCAL(p) // RC_NORM(p) +#define RC_NORM_REMOTE(p) RC_NORM(p) + +#define R (&p->rc.dec) + +MY_FORCE_INLINE +// MY_NO_INLINE +static void RangeDec_Decode(CPpmd7 *p, UInt32 start, UInt32 size) +{ + + + R->Code -= start * R->Range; + R->Range *= size; + RC_NORM_LOCAL(R) +} + +#define RC_Decode(start, size) RangeDec_Decode(p, start, size); +#define RC_DecodeFinal(start, size) RC_Decode(start, size) RC_NORM_REMOTE(R) +#define RC_GetThreshold(total) (R->Code / (R->Range /= (total))) + + +#define CTX(ref) ((CPpmd7_Context *)Ppmd7_GetContext(p, ref)) +typedef CPpmd7_Context * CTX_PTR; +#define SUCCESSOR(p) Ppmd_GET_SUCCESSOR(p) +void Ppmd7_UpdateModel(CPpmd7 *p); + +#define MASK(sym) ((unsigned char *)charMask)[sym] +// MY_FORCE_INLINE +// static +int Ppmd7z_DecodeSymbol(CPpmd7 *p) +{ + size_t charMask[256 / sizeof(size_t)]; + + if (p->MinContext->NumStats != 1) + { + CPpmd_State *s = Ppmd7_GetStats(p, p->MinContext); + unsigned i; + UInt32 count, hiCnt; + UInt32 summFreq = p->MinContext->Union2.SummFreq; + + + + + count = RC_GetThreshold(summFreq); + hiCnt = count; + + if ((Int32)(count -= s->Freq) < 0) + { + Byte sym; + RC_DecodeFinal(0, s->Freq); + p->FoundState = s; + sym = s->Symbol; + Ppmd7_Update1_0(p); + return sym; + } + + p->PrevSuccess = 0; + i = (unsigned)p->MinContext->NumStats - 1; + + do + { + if ((Int32)(count -= (++s)->Freq) < 0) + { + Byte sym; + RC_DecodeFinal((hiCnt - count) - s->Freq, s->Freq); + p->FoundState = s; + sym = s->Symbol; + Ppmd7_Update1(p); + return sym; + } + } + while (--i); + + if (hiCnt >= summFreq) + return PPMD7_SYM_ERROR; + + hiCnt -= count; + RC_Decode(hiCnt, summFreq - hiCnt); + + p->HiBitsFlag = PPMD7_HiBitsFlag_3(p->FoundState->Symbol); + PPMD_SetAllBitsIn256Bytes(charMask); + // i = p->MinContext->NumStats - 1; + // do { MASK((--s)->Symbol) = 0; } while (--i); + { + CPpmd_State *s2 = Ppmd7_GetStats(p, p->MinContext); + MASK(s->Symbol) = 0; + do + { + unsigned sym0 = s2[0].Symbol; + unsigned sym1 = s2[1].Symbol; + s2 += 2; + MASK(sym0) = 0; + MASK(sym1) = 0; + } + while (s2 < s); + } + } + else + { + CPpmd_State *s = Ppmd7Context_OneState(p->MinContext); + UInt16 *prob = Ppmd7_GetBinSumm(p); + UInt32 pr = *prob; + UInt32 size0 = (R->Range >> 14) * pr; + pr = PPMD_UPDATE_PROB_1(pr); + + if (R->Code < size0) + { + Byte sym; + *prob = (UInt16)(pr + (1 << PPMD_INT_BITS)); + + // RangeDec_DecodeBit0(size0); + R->Range = size0; + RC_NORM_1(R) + /* we can use single byte normalization here because of + (min(BinSumm[][]) = 95) > (1 << (14 - 8)) */ + + // sym = (p->FoundState = Ppmd7Context_OneState(p->MinContext))->Symbol; + // Ppmd7_UpdateBin(p); + { + unsigned freq = s->Freq; + CTX_PTR c = CTX(SUCCESSOR(s)); + sym = s->Symbol; + p->FoundState = s; + p->PrevSuccess = 1; + p->RunLength++; + s->Freq = (Byte)(freq + (freq < 128)); + // NextContext(p); + if (p->OrderFall == 0 && (const Byte *)c > p->Text) + p->MaxContext = p->MinContext = c; + else + Ppmd7_UpdateModel(p); + } + return sym; + } + + *prob = (UInt16)pr; + p->InitEsc = p->ExpEscape[pr >> 10]; + + // RangeDec_DecodeBit1(size0); + + R->Code -= size0; + R->Range -= size0; + RC_NORM_LOCAL(R) + + PPMD_SetAllBitsIn256Bytes(charMask); + MASK(Ppmd7Context_OneState(p->MinContext)->Symbol) = 0; + p->PrevSuccess = 0; + } + + for (;;) + { + CPpmd_State *s, *s2; + UInt32 freqSum, count, hiCnt; + + CPpmd_See *see; + CPpmd7_Context *mc; + unsigned numMasked; + RC_NORM_REMOTE(R) + mc = p->MinContext; + numMasked = mc->NumStats; + + do + { + p->OrderFall++; + if (!mc->Suffix) + return PPMD7_SYM_END; + mc = Ppmd7_GetContext(p, mc->Suffix); + } + while (mc->NumStats == numMasked); + + s = Ppmd7_GetStats(p, mc); + + { + unsigned num = mc->NumStats; + unsigned num2 = num / 2; + + num &= 1; + hiCnt = (s->Freq & (unsigned)(MASK(s->Symbol))) & (0 - (UInt32)num); + s += num; + p->MinContext = mc; + + do + { + unsigned sym0 = s[0].Symbol; + unsigned sym1 = s[1].Symbol; + s += 2; + hiCnt += (s[-2].Freq & (unsigned)(MASK(sym0))); + hiCnt += (s[-1].Freq & (unsigned)(MASK(sym1))); + } + while (--num2); + } + + see = Ppmd7_MakeEscFreq(p, numMasked, &freqSum); + freqSum += hiCnt; + + + + + count = RC_GetThreshold(freqSum); + + if (count < hiCnt) + { + Byte sym; + + s = Ppmd7_GetStats(p, p->MinContext); + hiCnt = count; + // count -= s->Freq & (unsigned)(MASK(s->Symbol)); + // if ((Int32)count >= 0) + { + for (;;) + { + count -= s->Freq & (unsigned)(MASK((s)->Symbol)); s++; if ((Int32)count < 0) break; + // count -= s->Freq & (unsigned)(MASK((s)->Symbol)); s++; if ((Int32)count < 0) break; + }; + } + s--; + RC_DecodeFinal((hiCnt - count) - s->Freq, s->Freq); + + // new (see->Summ) value can overflow over 16-bits in some rare cases + Ppmd_See_Update(see); + p->FoundState = s; + sym = s->Symbol; + Ppmd7_Update2(p); + return sym; + } + + if (count >= freqSum) + return PPMD7_SYM_ERROR; + + RC_Decode(hiCnt, freqSum - hiCnt); + + // We increase (see->Summ) for sum of Freqs of all non_Masked symbols. + // new (see->Summ) value can overflow over 16-bits in some rare cases + see->Summ = (UInt16)(see->Summ + freqSum); + + s = Ppmd7_GetStats(p, p->MinContext); + s2 = s + p->MinContext->NumStats; + do + { + MASK(s->Symbol) = 0; + s++; + } + while (s != s2); + } +} + +/* +Byte *Ppmd7z_DecodeSymbols(CPpmd7 *p, Byte *buf, const Byte *lim) +{ + int sym = 0; + if (buf != lim) + do + { + sym = Ppmd7z_DecodeSymbol(p); + if (sym < 0) + break; + *buf = (Byte)sym; + } + while (++buf < lim); + p->LastSymbol = sym; + return buf; +} +*/ diff --git a/deps/LZMA-SDK/C/Ppmd7Enc.c b/deps/LZMA-SDK/C/Ppmd7Enc.c index 6af1ec15e..62139c5b6 100644 --- a/deps/LZMA-SDK/C/Ppmd7Enc.c +++ b/deps/LZMA-SDK/C/Ppmd7Enc.c @@ -1,323 +1,323 @@ -/* Ppmd7Enc.c -- Ppmd7z (PPMdH with 7z Range Coder) Encoder -2021-04-13 : Igor Pavlov : Public domain -This code is based on: - PPMd var.H (2001): Dmitry Shkarin : Public domain */ - - -#include "Precomp.h" - -#include "Ppmd7.h" - -#define kTopValue (1 << 24) - -#define R (&p->rc.enc) - -void Ppmd7z_Init_RangeEnc(CPpmd7 *p) -{ - R->Low = 0; - R->Range = 0xFFFFFFFF; - R->Cache = 0; - R->CacheSize = 1; -} - -MY_NO_INLINE -static void RangeEnc_ShiftLow(CPpmd7 *p) -{ - if ((UInt32)R->Low < (UInt32)0xFF000000 || (unsigned)(R->Low >> 32) != 0) - { - Byte temp = R->Cache; - do - { - IByteOut_Write(R->Stream, (Byte)(temp + (Byte)(R->Low >> 32))); - temp = 0xFF; - } - while (--R->CacheSize != 0); - R->Cache = (Byte)((UInt32)R->Low >> 24); - } - R->CacheSize++; - R->Low = (UInt32)((UInt32)R->Low << 8); -} - -#define RC_NORM_BASE(p) if (R->Range < kTopValue) { R->Range <<= 8; RangeEnc_ShiftLow(p); -#define RC_NORM_1(p) RC_NORM_BASE(p) } -#define RC_NORM(p) RC_NORM_BASE(p) RC_NORM_BASE(p) }} - -// we must use only one type of Normalization from two: LOCAL or REMOTE -#define RC_NORM_LOCAL(p) // RC_NORM(p) -#define RC_NORM_REMOTE(p) RC_NORM(p) - -/* -#define RangeEnc_Encode(p, start, _size_) \ - { UInt32 size = _size_; \ - R->Low += start * R->Range; \ - R->Range *= size; \ - RC_NORM_LOCAL(p); } -*/ - -MY_FORCE_INLINE -// MY_NO_INLINE -static void RangeEnc_Encode(CPpmd7 *p, UInt32 start, UInt32 size) -{ - R->Low += start * R->Range; - R->Range *= size; - RC_NORM_LOCAL(p); -} - -void Ppmd7z_Flush_RangeEnc(CPpmd7 *p) -{ - unsigned i; - for (i = 0; i < 5; i++) - RangeEnc_ShiftLow(p); -} - - - -#define RC_Encode(start, size) RangeEnc_Encode(p, start, size); -#define RC_EncodeFinal(start, size) RC_Encode(start, size); RC_NORM_REMOTE(p); - -#define CTX(ref) ((CPpmd7_Context *)Ppmd7_GetContext(p, ref)) -#define SUFFIX(ctx) CTX((ctx)->Suffix) -typedef CPpmd7_Context * CTX_PTR; -#define SUCCESSOR(p) Ppmd_GET_SUCCESSOR(p) - -void Ppmd7_UpdateModel(CPpmd7 *p); - -#define MASK(sym) ((unsigned char *)charMask)[sym] - -MY_FORCE_INLINE -static -void Ppmd7z_EncodeSymbol(CPpmd7 *p, int symbol) -{ - size_t charMask[256 / sizeof(size_t)]; - - if (p->MinContext->NumStats != 1) - { - CPpmd_State *s = Ppmd7_GetStats(p, p->MinContext); - UInt32 sum; - unsigned i; - - - - - R->Range /= p->MinContext->Union2.SummFreq; - - if (s->Symbol == symbol) - { - // R->Range /= p->MinContext->Union2.SummFreq; - RC_EncodeFinal(0, s->Freq); - p->FoundState = s; - Ppmd7_Update1_0(p); - return; - } - p->PrevSuccess = 0; - sum = s->Freq; - i = (unsigned)p->MinContext->NumStats - 1; - do - { - if ((++s)->Symbol == symbol) - { - // R->Range /= p->MinContext->Union2.SummFreq; - RC_EncodeFinal(sum, s->Freq); - p->FoundState = s; - Ppmd7_Update1(p); - return; - } - sum += s->Freq; - } - while (--i); - - // R->Range /= p->MinContext->Union2.SummFreq; - RC_Encode(sum, p->MinContext->Union2.SummFreq - sum); - - p->HiBitsFlag = PPMD7_HiBitsFlag_3(p->FoundState->Symbol); - PPMD_SetAllBitsIn256Bytes(charMask); - // MASK(s->Symbol) = 0; - // i = p->MinContext->NumStats - 1; - // do { MASK((--s)->Symbol) = 0; } while (--i); - { - CPpmd_State *s2 = Ppmd7_GetStats(p, p->MinContext); - MASK(s->Symbol) = 0; - do - { - unsigned sym0 = s2[0].Symbol; - unsigned sym1 = s2[1].Symbol; - s2 += 2; - MASK(sym0) = 0; - MASK(sym1) = 0; - } - while (s2 < s); - } - } - else - { - UInt16 *prob = Ppmd7_GetBinSumm(p); - CPpmd_State *s = Ppmd7Context_OneState(p->MinContext); - UInt32 pr = *prob; - UInt32 bound = (R->Range >> 14) * pr; - pr = PPMD_UPDATE_PROB_1(pr); - if (s->Symbol == symbol) - { - *prob = (UInt16)(pr + (1 << PPMD_INT_BITS)); - // RangeEnc_EncodeBit_0(p, bound); - R->Range = bound; - RC_NORM_1(p); - - // p->FoundState = s; - // Ppmd7_UpdateBin(p); - { - unsigned freq = s->Freq; - CTX_PTR c = CTX(SUCCESSOR(s)); - p->FoundState = s; - p->PrevSuccess = 1; - p->RunLength++; - s->Freq = (Byte)(freq + (freq < 128)); - // NextContext(p); - if (p->OrderFall == 0 && (const Byte *)c > p->Text) - p->MaxContext = p->MinContext = c; - else - Ppmd7_UpdateModel(p); - } - return; - } - - *prob = (UInt16)pr; - p->InitEsc = p->ExpEscape[pr >> 10]; - // RangeEnc_EncodeBit_1(p, bound); - R->Low += bound; - R->Range -= bound; - RC_NORM_LOCAL(p) - - PPMD_SetAllBitsIn256Bytes(charMask); - MASK(s->Symbol) = 0; - p->PrevSuccess = 0; - } - - for (;;) - { - CPpmd_See *see; - CPpmd_State *s; - UInt32 sum, escFreq; - CPpmd7_Context *mc; - unsigned i, numMasked; - - RC_NORM_REMOTE(p) - - mc = p->MinContext; - numMasked = mc->NumStats; - - do - { - p->OrderFall++; - if (!mc->Suffix) - return; /* EndMarker (symbol = -1) */ - mc = Ppmd7_GetContext(p, mc->Suffix); - i = mc->NumStats; - } - while (i == numMasked); - - p->MinContext = mc; - - // see = Ppmd7_MakeEscFreq(p, numMasked, &escFreq); - { - if (i != 256) - { - unsigned nonMasked = i - numMasked; - see = p->See[(unsigned)p->NS2Indx[(size_t)nonMasked - 1]] - + p->HiBitsFlag - + (nonMasked < (unsigned)SUFFIX(mc)->NumStats - i) - + 2 * (unsigned)(mc->Union2.SummFreq < 11 * i) - + 4 * (unsigned)(numMasked > nonMasked); - { - // if (see->Summ) field is larger than 16-bit, we need only low 16 bits of Summ - unsigned summ = (UInt16)see->Summ; // & 0xFFFF - unsigned r = (summ >> see->Shift); - see->Summ = (UInt16)(summ - r); - escFreq = r + (r == 0); - } - } - else - { - see = &p->DummySee; - escFreq = 1; - } - } - - s = Ppmd7_GetStats(p, mc); - sum = 0; - // i = mc->NumStats; - - do - { - unsigned cur = s->Symbol; - if ((int)cur == symbol) - { - UInt32 low = sum; - UInt32 freq = s->Freq; - unsigned num2; - - Ppmd_See_Update(see); - p->FoundState = s; - sum += escFreq; - - num2 = i / 2; - i &= 1; - sum += freq & (0 - (UInt32)i); - if (num2 != 0) - { - s += i; - for (;;) - { - unsigned sym0 = s[0].Symbol; - unsigned sym1 = s[1].Symbol; - s += 2; - sum += (s[-2].Freq & (unsigned)(MASK(sym0))); - sum += (s[-1].Freq & (unsigned)(MASK(sym1))); - if (--num2 == 0) - break; - } - } - - - R->Range /= sum; - RC_EncodeFinal(low, freq); - Ppmd7_Update2(p); - return; - } - sum += (s->Freq & (unsigned)(MASK(cur))); - s++; - } - while (--i); - - { - UInt32 total = sum + escFreq; - see->Summ = (UInt16)(see->Summ + total); - - R->Range /= total; - RC_Encode(sum, escFreq); - } - - { - CPpmd_State *s2 = Ppmd7_GetStats(p, p->MinContext); - s--; - MASK(s->Symbol) = 0; - do - { - unsigned sym0 = s2[0].Symbol; - unsigned sym1 = s2[1].Symbol; - s2 += 2; - MASK(sym0) = 0; - MASK(sym1) = 0; - } - while (s2 < s); - } - } -} - - -void Ppmd7z_EncodeSymbols(CPpmd7 *p, const Byte *buf, const Byte *lim) -{ - for (; buf < lim; buf++) - { - Ppmd7z_EncodeSymbol(p, *buf); - } -} +/* Ppmd7Enc.c -- Ppmd7z (PPMdH with 7z Range Coder) Encoder +2021-04-13 : Igor Pavlov : Public domain +This code is based on: + PPMd var.H (2001): Dmitry Shkarin : Public domain */ + + +#include "Precomp.h" + +#include "Ppmd7.h" + +#define kTopValue (1 << 24) + +#define R (&p->rc.enc) + +void Ppmd7z_Init_RangeEnc(CPpmd7 *p) +{ + R->Low = 0; + R->Range = 0xFFFFFFFF; + R->Cache = 0; + R->CacheSize = 1; +} + +MY_NO_INLINE +static void RangeEnc_ShiftLow(CPpmd7 *p) +{ + if ((UInt32)R->Low < (UInt32)0xFF000000 || (unsigned)(R->Low >> 32) != 0) + { + Byte temp = R->Cache; + do + { + IByteOut_Write(R->Stream, (Byte)(temp + (Byte)(R->Low >> 32))); + temp = 0xFF; + } + while (--R->CacheSize != 0); + R->Cache = (Byte)((UInt32)R->Low >> 24); + } + R->CacheSize++; + R->Low = (UInt32)((UInt32)R->Low << 8); +} + +#define RC_NORM_BASE(p) if (R->Range < kTopValue) { R->Range <<= 8; RangeEnc_ShiftLow(p); +#define RC_NORM_1(p) RC_NORM_BASE(p) } +#define RC_NORM(p) RC_NORM_BASE(p) RC_NORM_BASE(p) }} + +// we must use only one type of Normalization from two: LOCAL or REMOTE +#define RC_NORM_LOCAL(p) // RC_NORM(p) +#define RC_NORM_REMOTE(p) RC_NORM(p) + +/* +#define RangeEnc_Encode(p, start, _size_) \ + { UInt32 size = _size_; \ + R->Low += start * R->Range; \ + R->Range *= size; \ + RC_NORM_LOCAL(p); } +*/ + +MY_FORCE_INLINE +// MY_NO_INLINE +static void RangeEnc_Encode(CPpmd7 *p, UInt32 start, UInt32 size) +{ + R->Low += start * R->Range; + R->Range *= size; + RC_NORM_LOCAL(p); +} + +void Ppmd7z_Flush_RangeEnc(CPpmd7 *p) +{ + unsigned i; + for (i = 0; i < 5; i++) + RangeEnc_ShiftLow(p); +} + + + +#define RC_Encode(start, size) RangeEnc_Encode(p, start, size); +#define RC_EncodeFinal(start, size) RC_Encode(start, size); RC_NORM_REMOTE(p); + +#define CTX(ref) ((CPpmd7_Context *)Ppmd7_GetContext(p, ref)) +#define SUFFIX(ctx) CTX((ctx)->Suffix) +typedef CPpmd7_Context * CTX_PTR; +#define SUCCESSOR(p) Ppmd_GET_SUCCESSOR(p) + +void Ppmd7_UpdateModel(CPpmd7 *p); + +#define MASK(sym) ((unsigned char *)charMask)[sym] + +MY_FORCE_INLINE +static +void Ppmd7z_EncodeSymbol(CPpmd7 *p, int symbol) +{ + size_t charMask[256 / sizeof(size_t)]; + + if (p->MinContext->NumStats != 1) + { + CPpmd_State *s = Ppmd7_GetStats(p, p->MinContext); + UInt32 sum; + unsigned i; + + + + + R->Range /= p->MinContext->Union2.SummFreq; + + if (s->Symbol == symbol) + { + // R->Range /= p->MinContext->Union2.SummFreq; + RC_EncodeFinal(0, s->Freq); + p->FoundState = s; + Ppmd7_Update1_0(p); + return; + } + p->PrevSuccess = 0; + sum = s->Freq; + i = (unsigned)p->MinContext->NumStats - 1; + do + { + if ((++s)->Symbol == symbol) + { + // R->Range /= p->MinContext->Union2.SummFreq; + RC_EncodeFinal(sum, s->Freq); + p->FoundState = s; + Ppmd7_Update1(p); + return; + } + sum += s->Freq; + } + while (--i); + + // R->Range /= p->MinContext->Union2.SummFreq; + RC_Encode(sum, p->MinContext->Union2.SummFreq - sum); + + p->HiBitsFlag = PPMD7_HiBitsFlag_3(p->FoundState->Symbol); + PPMD_SetAllBitsIn256Bytes(charMask); + // MASK(s->Symbol) = 0; + // i = p->MinContext->NumStats - 1; + // do { MASK((--s)->Symbol) = 0; } while (--i); + { + CPpmd_State *s2 = Ppmd7_GetStats(p, p->MinContext); + MASK(s->Symbol) = 0; + do + { + unsigned sym0 = s2[0].Symbol; + unsigned sym1 = s2[1].Symbol; + s2 += 2; + MASK(sym0) = 0; + MASK(sym1) = 0; + } + while (s2 < s); + } + } + else + { + UInt16 *prob = Ppmd7_GetBinSumm(p); + CPpmd_State *s = Ppmd7Context_OneState(p->MinContext); + UInt32 pr = *prob; + UInt32 bound = (R->Range >> 14) * pr; + pr = PPMD_UPDATE_PROB_1(pr); + if (s->Symbol == symbol) + { + *prob = (UInt16)(pr + (1 << PPMD_INT_BITS)); + // RangeEnc_EncodeBit_0(p, bound); + R->Range = bound; + RC_NORM_1(p); + + // p->FoundState = s; + // Ppmd7_UpdateBin(p); + { + unsigned freq = s->Freq; + CTX_PTR c = CTX(SUCCESSOR(s)); + p->FoundState = s; + p->PrevSuccess = 1; + p->RunLength++; + s->Freq = (Byte)(freq + (freq < 128)); + // NextContext(p); + if (p->OrderFall == 0 && (const Byte *)c > p->Text) + p->MaxContext = p->MinContext = c; + else + Ppmd7_UpdateModel(p); + } + return; + } + + *prob = (UInt16)pr; + p->InitEsc = p->ExpEscape[pr >> 10]; + // RangeEnc_EncodeBit_1(p, bound); + R->Low += bound; + R->Range -= bound; + RC_NORM_LOCAL(p) + + PPMD_SetAllBitsIn256Bytes(charMask); + MASK(s->Symbol) = 0; + p->PrevSuccess = 0; + } + + for (;;) + { + CPpmd_See *see; + CPpmd_State *s; + UInt32 sum, escFreq; + CPpmd7_Context *mc; + unsigned i, numMasked; + + RC_NORM_REMOTE(p) + + mc = p->MinContext; + numMasked = mc->NumStats; + + do + { + p->OrderFall++; + if (!mc->Suffix) + return; /* EndMarker (symbol = -1) */ + mc = Ppmd7_GetContext(p, mc->Suffix); + i = mc->NumStats; + } + while (i == numMasked); + + p->MinContext = mc; + + // see = Ppmd7_MakeEscFreq(p, numMasked, &escFreq); + { + if (i != 256) + { + unsigned nonMasked = i - numMasked; + see = p->See[(unsigned)p->NS2Indx[(size_t)nonMasked - 1]] + + p->HiBitsFlag + + (nonMasked < (unsigned)SUFFIX(mc)->NumStats - i) + + 2 * (unsigned)(mc->Union2.SummFreq < 11 * i) + + 4 * (unsigned)(numMasked > nonMasked); + { + // if (see->Summ) field is larger than 16-bit, we need only low 16 bits of Summ + unsigned summ = (UInt16)see->Summ; // & 0xFFFF + unsigned r = (summ >> see->Shift); + see->Summ = (UInt16)(summ - r); + escFreq = r + (r == 0); + } + } + else + { + see = &p->DummySee; + escFreq = 1; + } + } + + s = Ppmd7_GetStats(p, mc); + sum = 0; + // i = mc->NumStats; + + do + { + unsigned cur = s->Symbol; + if ((int)cur == symbol) + { + UInt32 low = sum; + UInt32 freq = s->Freq; + unsigned num2; + + Ppmd_See_Update(see); + p->FoundState = s; + sum += escFreq; + + num2 = i / 2; + i &= 1; + sum += freq & (0 - (UInt32)i); + if (num2 != 0) + { + s += i; + for (;;) + { + unsigned sym0 = s[0].Symbol; + unsigned sym1 = s[1].Symbol; + s += 2; + sum += (s[-2].Freq & (unsigned)(MASK(sym0))); + sum += (s[-1].Freq & (unsigned)(MASK(sym1))); + if (--num2 == 0) + break; + } + } + + + R->Range /= sum; + RC_EncodeFinal(low, freq); + Ppmd7_Update2(p); + return; + } + sum += (s->Freq & (unsigned)(MASK(cur))); + s++; + } + while (--i); + + { + UInt32 total = sum + escFreq; + see->Summ = (UInt16)(see->Summ + total); + + R->Range /= total; + RC_Encode(sum, escFreq); + } + + { + CPpmd_State *s2 = Ppmd7_GetStats(p, p->MinContext); + s--; + MASK(s->Symbol) = 0; + do + { + unsigned sym0 = s2[0].Symbol; + unsigned sym1 = s2[1].Symbol; + s2 += 2; + MASK(sym0) = 0; + MASK(sym1) = 0; + } + while (s2 < s); + } + } +} + + +void Ppmd7z_EncodeSymbols(CPpmd7 *p, const Byte *buf, const Byte *lim) +{ + for (; buf < lim; buf++) + { + Ppmd7z_EncodeSymbol(p, *buf); + } +} diff --git a/deps/LZMA-SDK/C/Precomp.h b/deps/LZMA-SDK/C/Precomp.h index edb581443..e8ff8b40e 100644 --- a/deps/LZMA-SDK/C/Precomp.h +++ b/deps/LZMA-SDK/C/Precomp.h @@ -1,10 +1,10 @@ -/* Precomp.h -- StdAfx -2013-11-12 : Igor Pavlov : Public domain */ - -#ifndef __7Z_PRECOMP_H -#define __7Z_PRECOMP_H - -#include "Compiler.h" -/* #include "7zTypes.h" */ - -#endif +/* Precomp.h -- StdAfx +2013-11-12 : Igor Pavlov : Public domain */ + +#ifndef __7Z_PRECOMP_H +#define __7Z_PRECOMP_H + +#include "Compiler.h" +/* #include "7zTypes.h" */ + +#endif diff --git a/deps/LZMA-SDK/C/RotateDefs.h b/deps/LZMA-SDK/C/RotateDefs.h index 6c790e791..8f01d1a6c 100644 --- a/deps/LZMA-SDK/C/RotateDefs.h +++ b/deps/LZMA-SDK/C/RotateDefs.h @@ -1,30 +1,30 @@ -/* RotateDefs.h -- Rotate functions -2015-03-25 : Igor Pavlov : Public domain */ - -#ifndef __ROTATE_DEFS_H -#define __ROTATE_DEFS_H - -#ifdef _MSC_VER - -#include - -/* don't use _rotl with MINGW. It can insert slow call to function. */ - -/* #if (_MSC_VER >= 1200) */ -#pragma intrinsic(_rotl) -#pragma intrinsic(_rotr) -/* #endif */ - -#define rotlFixed(x, n) _rotl((x), (n)) -#define rotrFixed(x, n) _rotr((x), (n)) - -#else - -/* new compilers can translate these macros to fast commands. */ - -#define rotlFixed(x, n) (((x) << (n)) | ((x) >> (32 - (n)))) -#define rotrFixed(x, n) (((x) >> (n)) | ((x) << (32 - (n)))) - -#endif - -#endif +/* RotateDefs.h -- Rotate functions +2015-03-25 : Igor Pavlov : Public domain */ + +#ifndef __ROTATE_DEFS_H +#define __ROTATE_DEFS_H + +#ifdef _MSC_VER + +#include + +/* don't use _rotl with MINGW. It can insert slow call to function. */ + +/* #if (_MSC_VER >= 1200) */ +#pragma intrinsic(_rotl) +#pragma intrinsic(_rotr) +/* #endif */ + +#define rotlFixed(x, n) _rotl((x), (n)) +#define rotrFixed(x, n) _rotr((x), (n)) + +#else + +/* new compilers can translate these macros to fast commands. */ + +#define rotlFixed(x, n) (((x) << (n)) | ((x) >> (32 - (n)))) +#define rotrFixed(x, n) (((x) >> (n)) | ((x) << (32 - (n)))) + +#endif + +#endif diff --git a/deps/LZMA-SDK/C/Sha256.c b/deps/LZMA-SDK/C/Sha256.c index c03b75afe..8b3983ea7 100644 --- a/deps/LZMA-SDK/C/Sha256.c +++ b/deps/LZMA-SDK/C/Sha256.c @@ -1,486 +1,486 @@ -/* Sha256.c -- SHA-256 Hash -2021-04-01 : Igor Pavlov : Public domain -This code is based on public domain code from Wei Dai's Crypto++ library. */ - -#include "Precomp.h" - -#include - -#include "CpuArch.h" -#include "RotateDefs.h" -#include "Sha256.h" - -#if defined(_MSC_VER) && (_MSC_VER < 1900) -// #define USE_MY_MM -#endif - -#ifdef MY_CPU_X86_OR_AMD64 - #ifdef _MSC_VER - #if _MSC_VER >= 1200 - #define _SHA_SUPPORTED - #endif - #elif defined(__clang__) - #if (__clang_major__ >= 8) // fix that check - #define _SHA_SUPPORTED - #endif - #elif defined(__GNUC__) - #if (__GNUC__ >= 8) // fix that check - #define _SHA_SUPPORTED - #endif - #elif defined(__INTEL_COMPILER) - #if (__INTEL_COMPILER >= 1800) // fix that check - #define _SHA_SUPPORTED - #endif - #endif -#elif defined(MY_CPU_ARM_OR_ARM64) - #ifdef _MSC_VER - #if _MSC_VER >= 1910 - #define _SHA_SUPPORTED - #endif - #elif defined(__clang__) - #if (__clang_major__ >= 8) // fix that check - #define _SHA_SUPPORTED - #endif - #elif defined(__GNUC__) - #if (__GNUC__ >= 6) // fix that check - #define _SHA_SUPPORTED - #endif - #endif -#endif - -void MY_FAST_CALL Sha256_UpdateBlocks(UInt32 state[8], const Byte *data, size_t numBlocks); - -#ifdef _SHA_SUPPORTED - void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); - - static SHA256_FUNC_UPDATE_BLOCKS g_FUNC_UPDATE_BLOCKS = Sha256_UpdateBlocks; - static SHA256_FUNC_UPDATE_BLOCKS g_FUNC_UPDATE_BLOCKS_HW; - - #define UPDATE_BLOCKS(p) p->func_UpdateBlocks -#else - #define UPDATE_BLOCKS(p) Sha256_UpdateBlocks -#endif - - -BoolInt Sha256_SetFunction(CSha256 *p, unsigned algo) -{ - SHA256_FUNC_UPDATE_BLOCKS func = Sha256_UpdateBlocks; - - #ifdef _SHA_SUPPORTED - if (algo != SHA256_ALGO_SW) - { - if (algo == SHA256_ALGO_DEFAULT) - func = g_FUNC_UPDATE_BLOCKS; - else - { - if (algo != SHA256_ALGO_HW) - return False; - func = g_FUNC_UPDATE_BLOCKS_HW; - if (!func) - return False; - } - } - #else - if (algo > 1) - return False; - #endif - - p->func_UpdateBlocks = func; - return True; -} - - -/* define it for speed optimization */ - -#ifdef _SFX - #define STEP_PRE 1 - #define STEP_MAIN 1 -#else - #define STEP_PRE 2 - #define STEP_MAIN 4 - // #define _SHA256_UNROLL -#endif - -#if STEP_MAIN != 16 - #define _SHA256_BIG_W -#endif - - - - -void Sha256_InitState(CSha256 *p) -{ - p->count = 0; - p->state[0] = 0x6a09e667; - p->state[1] = 0xbb67ae85; - p->state[2] = 0x3c6ef372; - p->state[3] = 0xa54ff53a; - p->state[4] = 0x510e527f; - p->state[5] = 0x9b05688c; - p->state[6] = 0x1f83d9ab; - p->state[7] = 0x5be0cd19; -} - -void Sha256_Init(CSha256 *p) -{ - p->func_UpdateBlocks = - #ifdef _SHA_SUPPORTED - g_FUNC_UPDATE_BLOCKS; - #else - NULL; - #endif - Sha256_InitState(p); -} - -#define S0(x) (rotrFixed(x, 2) ^ rotrFixed(x,13) ^ rotrFixed(x, 22)) -#define S1(x) (rotrFixed(x, 6) ^ rotrFixed(x,11) ^ rotrFixed(x, 25)) -#define s0(x) (rotrFixed(x, 7) ^ rotrFixed(x,18) ^ (x >> 3)) -#define s1(x) (rotrFixed(x,17) ^ rotrFixed(x,19) ^ (x >> 10)) - -#define Ch(x,y,z) (z^(x&(y^z))) -#define Maj(x,y,z) ((x&y)|(z&(x|y))) - - -#define W_PRE(i) (W[(i) + (size_t)(j)] = GetBe32(data + ((size_t)(j) + i) * 4)) - -#define blk2_main(j, i) s1(w(j, (i)-2)) + w(j, (i)-7) + s0(w(j, (i)-15)) - -#ifdef _SHA256_BIG_W - // we use +i instead of +(i) to change the order to solve CLANG compiler warning for signed/unsigned. - #define w(j, i) W[(size_t)(j) + i] - #define blk2(j, i) (w(j, i) = w(j, (i)-16) + blk2_main(j, i)) -#else - #if STEP_MAIN == 16 - #define w(j, i) W[(i) & 15] - #else - #define w(j, i) W[((size_t)(j) + (i)) & 15] - #endif - #define blk2(j, i) (w(j, i) += blk2_main(j, i)) -#endif - -#define W_MAIN(i) blk2(j, i) - - -#define T1(wx, i) \ - tmp = h + S1(e) + Ch(e,f,g) + K[(i)+(size_t)(j)] + wx(i); \ - h = g; \ - g = f; \ - f = e; \ - e = d + tmp; \ - tmp += S0(a) + Maj(a, b, c); \ - d = c; \ - c = b; \ - b = a; \ - a = tmp; \ - -#define R1_PRE(i) T1( W_PRE, i) -#define R1_MAIN(i) T1( W_MAIN, i) - -#if (!defined(_SHA256_UNROLL) || STEP_MAIN < 8) && (STEP_MAIN >= 4) -#define R2_MAIN(i) \ - R1_MAIN(i) \ - R1_MAIN(i + 1) \ - -#endif - - - -#if defined(_SHA256_UNROLL) && STEP_MAIN >= 8 - -#define T4( a,b,c,d,e,f,g,h, wx, i) \ - h += S1(e) + Ch(e,f,g) + K[(i)+(size_t)(j)] + wx(i); \ - tmp = h; \ - h += d; \ - d = tmp + S0(a) + Maj(a, b, c); \ - -#define R4( wx, i) \ - T4 ( a,b,c,d,e,f,g,h, wx, (i )); \ - T4 ( d,a,b,c,h,e,f,g, wx, (i+1)); \ - T4 ( c,d,a,b,g,h,e,f, wx, (i+2)); \ - T4 ( b,c,d,a,f,g,h,e, wx, (i+3)); \ - -#define R4_PRE(i) R4( W_PRE, i) -#define R4_MAIN(i) R4( W_MAIN, i) - - -#define T8( a,b,c,d,e,f,g,h, wx, i) \ - h += S1(e) + Ch(e,f,g) + K[(i)+(size_t)(j)] + wx(i); \ - d += h; \ - h += S0(a) + Maj(a, b, c); \ - -#define R8( wx, i) \ - T8 ( a,b,c,d,e,f,g,h, wx, i ); \ - T8 ( h,a,b,c,d,e,f,g, wx, i+1); \ - T8 ( g,h,a,b,c,d,e,f, wx, i+2); \ - T8 ( f,g,h,a,b,c,d,e, wx, i+3); \ - T8 ( e,f,g,h,a,b,c,d, wx, i+4); \ - T8 ( d,e,f,g,h,a,b,c, wx, i+5); \ - T8 ( c,d,e,f,g,h,a,b, wx, i+6); \ - T8 ( b,c,d,e,f,g,h,a, wx, i+7); \ - -#define R8_PRE(i) R8( W_PRE, i) -#define R8_MAIN(i) R8( W_MAIN, i) - -#endif - -void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); - -// static -extern MY_ALIGN(64) -const UInt32 SHA256_K_ARRAY[64]; - -MY_ALIGN(64) -const UInt32 SHA256_K_ARRAY[64] = { - 0x428a2f98, 0x71374491, 0xb5c0fbcf, 0xe9b5dba5, - 0x3956c25b, 0x59f111f1, 0x923f82a4, 0xab1c5ed5, - 0xd807aa98, 0x12835b01, 0x243185be, 0x550c7dc3, - 0x72be5d74, 0x80deb1fe, 0x9bdc06a7, 0xc19bf174, - 0xe49b69c1, 0xefbe4786, 0x0fc19dc6, 0x240ca1cc, - 0x2de92c6f, 0x4a7484aa, 0x5cb0a9dc, 0x76f988da, - 0x983e5152, 0xa831c66d, 0xb00327c8, 0xbf597fc7, - 0xc6e00bf3, 0xd5a79147, 0x06ca6351, 0x14292967, - 0x27b70a85, 0x2e1b2138, 0x4d2c6dfc, 0x53380d13, - 0x650a7354, 0x766a0abb, 0x81c2c92e, 0x92722c85, - 0xa2bfe8a1, 0xa81a664b, 0xc24b8b70, 0xc76c51a3, - 0xd192e819, 0xd6990624, 0xf40e3585, 0x106aa070, - 0x19a4c116, 0x1e376c08, 0x2748774c, 0x34b0bcb5, - 0x391c0cb3, 0x4ed8aa4a, 0x5b9cca4f, 0x682e6ff3, - 0x748f82ee, 0x78a5636f, 0x84c87814, 0x8cc70208, - 0x90befffa, 0xa4506ceb, 0xbef9a3f7, 0xc67178f2 -}; - -#define K SHA256_K_ARRAY - - -MY_NO_INLINE -void MY_FAST_CALL Sha256_UpdateBlocks(UInt32 state[8], const Byte *data, size_t numBlocks) -{ - UInt32 W - #ifdef _SHA256_BIG_W - [64]; - #else - [16]; - #endif - - unsigned j; - - UInt32 a,b,c,d,e,f,g,h; - - #if !defined(_SHA256_UNROLL) || (STEP_MAIN <= 4) || (STEP_PRE <= 4) - UInt32 tmp; - #endif - - a = state[0]; - b = state[1]; - c = state[2]; - d = state[3]; - e = state[4]; - f = state[5]; - g = state[6]; - h = state[7]; - - while (numBlocks) - { - - for (j = 0; j < 16; j += STEP_PRE) - { - #if STEP_PRE > 4 - - #if STEP_PRE < 8 - R4_PRE(0); - #else - R8_PRE(0); - #if STEP_PRE == 16 - R8_PRE(8); - #endif - #endif - - #else - - R1_PRE(0); - #if STEP_PRE >= 2 - R1_PRE(1); - #if STEP_PRE >= 4 - R1_PRE(2); - R1_PRE(3); - #endif - #endif - - #endif - } - - for (j = 16; j < 64; j += STEP_MAIN) - { - #if defined(_SHA256_UNROLL) && STEP_MAIN >= 8 - - #if STEP_MAIN < 8 - R4_MAIN(0); - #else - R8_MAIN(0); - #if STEP_MAIN == 16 - R8_MAIN(8); - #endif - #endif - - #else - - R1_MAIN(0); - #if STEP_MAIN >= 2 - R1_MAIN(1); - #if STEP_MAIN >= 4 - R2_MAIN(2); - #if STEP_MAIN >= 8 - R2_MAIN(4); - R2_MAIN(6); - #if STEP_MAIN >= 16 - R2_MAIN(8); - R2_MAIN(10); - R2_MAIN(12); - R2_MAIN(14); - #endif - #endif - #endif - #endif - #endif - } - - a += state[0]; state[0] = a; - b += state[1]; state[1] = b; - c += state[2]; state[2] = c; - d += state[3]; state[3] = d; - e += state[4]; state[4] = e; - f += state[5]; state[5] = f; - g += state[6]; state[6] = g; - h += state[7]; state[7] = h; - - data += 64; - numBlocks--; - } - - /* Wipe variables */ - /* memset(W, 0, sizeof(W)); */ -} - -#undef S0 -#undef S1 -#undef s0 -#undef s1 -#undef K - -#define Sha256_UpdateBlock(p) UPDATE_BLOCKS(p)(p->state, p->buffer, 1) - -void Sha256_Update(CSha256 *p, const Byte *data, size_t size) -{ - if (size == 0) - return; - - { - unsigned pos = (unsigned)p->count & 0x3F; - unsigned num; - - p->count += size; - - num = 64 - pos; - if (num > size) - { - memcpy(p->buffer + pos, data, size); - return; - } - - if (pos != 0) - { - size -= num; - memcpy(p->buffer + pos, data, num); - data += num; - Sha256_UpdateBlock(p); - } - } - { - size_t numBlocks = size >> 6; - UPDATE_BLOCKS(p)(p->state, data, numBlocks); - size &= 0x3F; - if (size == 0) - return; - data += (numBlocks << 6); - memcpy(p->buffer, data, size); - } -} - - -void Sha256_Final(CSha256 *p, Byte *digest) -{ - unsigned pos = (unsigned)p->count & 0x3F; - unsigned i; - - p->buffer[pos++] = 0x80; - - if (pos > (64 - 8)) - { - while (pos != 64) { p->buffer[pos++] = 0; } - // memset(&p->buf.buffer[pos], 0, 64 - pos); - Sha256_UpdateBlock(p); - pos = 0; - } - - /* - if (pos & 3) - { - p->buffer[pos] = 0; - p->buffer[pos + 1] = 0; - p->buffer[pos + 2] = 0; - pos += 3; - pos &= ~3; - } - { - for (; pos < 64 - 8; pos += 4) - *(UInt32 *)(&p->buffer[pos]) = 0; - } - */ - - memset(&p->buffer[pos], 0, (64 - 8) - pos); - - { - UInt64 numBits = (p->count << 3); - SetBe32(p->buffer + 64 - 8, (UInt32)(numBits >> 32)); - SetBe32(p->buffer + 64 - 4, (UInt32)(numBits)); - } - - Sha256_UpdateBlock(p); - - for (i = 0; i < 8; i += 2) - { - UInt32 v0 = p->state[i]; - UInt32 v1 = p->state[(size_t)i + 1]; - SetBe32(digest , v0); - SetBe32(digest + 4, v1); - digest += 8; - } - - Sha256_InitState(p); -} - - -void Sha256Prepare() -{ - #ifdef _SHA_SUPPORTED - SHA256_FUNC_UPDATE_BLOCKS f, f_hw; - f = Sha256_UpdateBlocks; - f_hw = NULL; - #ifdef MY_CPU_X86_OR_AMD64 - #ifndef USE_MY_MM - if (CPU_IsSupported_SHA() - && CPU_IsSupported_SSSE3() - // && CPU_IsSupported_SSE41() - ) - #endif - #else - if (CPU_IsSupported_SHA2()) - #endif - { - // printf("\n========== HW SHA256 ======== \n"); - f = f_hw = Sha256_UpdateBlocks_HW; - } - g_FUNC_UPDATE_BLOCKS = f; - g_FUNC_UPDATE_BLOCKS_HW = f_hw; - #endif -} +/* Sha256.c -- SHA-256 Hash +2021-04-01 : Igor Pavlov : Public domain +This code is based on public domain code from Wei Dai's Crypto++ library. */ + +#include "Precomp.h" + +#include + +#include "CpuArch.h" +#include "RotateDefs.h" +#include "Sha256.h" + +#if defined(_MSC_VER) && (_MSC_VER < 1900) +// #define USE_MY_MM +#endif + +#ifdef MY_CPU_X86_OR_AMD64 + #ifdef _MSC_VER + #if _MSC_VER >= 1200 + #define _SHA_SUPPORTED + #endif + #elif defined(__clang__) + #if (__clang_major__ >= 8) // fix that check + #define _SHA_SUPPORTED + #endif + #elif defined(__GNUC__) + #if (__GNUC__ >= 8) // fix that check + #define _SHA_SUPPORTED + #endif + #elif defined(__INTEL_COMPILER) + #if (__INTEL_COMPILER >= 1800) // fix that check + #define _SHA_SUPPORTED + #endif + #endif +#elif defined(MY_CPU_ARM_OR_ARM64) + #ifdef _MSC_VER + #if _MSC_VER >= 1910 + #define _SHA_SUPPORTED + #endif + #elif defined(__clang__) + #if (__clang_major__ >= 8) // fix that check + #define _SHA_SUPPORTED + #endif + #elif defined(__GNUC__) + #if (__GNUC__ >= 6) // fix that check + #define _SHA_SUPPORTED + #endif + #endif +#endif + +void MY_FAST_CALL Sha256_UpdateBlocks(UInt32 state[8], const Byte *data, size_t numBlocks); + +#ifdef _SHA_SUPPORTED + void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); + + static SHA256_FUNC_UPDATE_BLOCKS g_FUNC_UPDATE_BLOCKS = Sha256_UpdateBlocks; + static SHA256_FUNC_UPDATE_BLOCKS g_FUNC_UPDATE_BLOCKS_HW; + + #define UPDATE_BLOCKS(p) p->func_UpdateBlocks +#else + #define UPDATE_BLOCKS(p) Sha256_UpdateBlocks +#endif + + +BoolInt Sha256_SetFunction(CSha256 *p, unsigned algo) +{ + SHA256_FUNC_UPDATE_BLOCKS func = Sha256_UpdateBlocks; + + #ifdef _SHA_SUPPORTED + if (algo != SHA256_ALGO_SW) + { + if (algo == SHA256_ALGO_DEFAULT) + func = g_FUNC_UPDATE_BLOCKS; + else + { + if (algo != SHA256_ALGO_HW) + return False; + func = g_FUNC_UPDATE_BLOCKS_HW; + if (!func) + return False; + } + } + #else + if (algo > 1) + return False; + #endif + + p->func_UpdateBlocks = func; + return True; +} + + +/* define it for speed optimization */ + +#ifdef _SFX + #define STEP_PRE 1 + #define STEP_MAIN 1 +#else + #define STEP_PRE 2 + #define STEP_MAIN 4 + // #define _SHA256_UNROLL +#endif + +#if STEP_MAIN != 16 + #define _SHA256_BIG_W +#endif + + + + +void Sha256_InitState(CSha256 *p) +{ + p->count = 0; + p->state[0] = 0x6a09e667; + p->state[1] = 0xbb67ae85; + p->state[2] = 0x3c6ef372; + p->state[3] = 0xa54ff53a; + p->state[4] = 0x510e527f; + p->state[5] = 0x9b05688c; + p->state[6] = 0x1f83d9ab; + p->state[7] = 0x5be0cd19; +} + +void Sha256_Init(CSha256 *p) +{ + p->func_UpdateBlocks = + #ifdef _SHA_SUPPORTED + g_FUNC_UPDATE_BLOCKS; + #else + NULL; + #endif + Sha256_InitState(p); +} + +#define S0(x) (rotrFixed(x, 2) ^ rotrFixed(x,13) ^ rotrFixed(x, 22)) +#define S1(x) (rotrFixed(x, 6) ^ rotrFixed(x,11) ^ rotrFixed(x, 25)) +#define s0(x) (rotrFixed(x, 7) ^ rotrFixed(x,18) ^ (x >> 3)) +#define s1(x) (rotrFixed(x,17) ^ rotrFixed(x,19) ^ (x >> 10)) + +#define Ch(x,y,z) (z^(x&(y^z))) +#define Maj(x,y,z) ((x&y)|(z&(x|y))) + + +#define W_PRE(i) (W[(i) + (size_t)(j)] = GetBe32(data + ((size_t)(j) + i) * 4)) + +#define blk2_main(j, i) s1(w(j, (i)-2)) + w(j, (i)-7) + s0(w(j, (i)-15)) + +#ifdef _SHA256_BIG_W + // we use +i instead of +(i) to change the order to solve CLANG compiler warning for signed/unsigned. + #define w(j, i) W[(size_t)(j) + i] + #define blk2(j, i) (w(j, i) = w(j, (i)-16) + blk2_main(j, i)) +#else + #if STEP_MAIN == 16 + #define w(j, i) W[(i) & 15] + #else + #define w(j, i) W[((size_t)(j) + (i)) & 15] + #endif + #define blk2(j, i) (w(j, i) += blk2_main(j, i)) +#endif + +#define W_MAIN(i) blk2(j, i) + + +#define T1(wx, i) \ + tmp = h + S1(e) + Ch(e,f,g) + K[(i)+(size_t)(j)] + wx(i); \ + h = g; \ + g = f; \ + f = e; \ + e = d + tmp; \ + tmp += S0(a) + Maj(a, b, c); \ + d = c; \ + c = b; \ + b = a; \ + a = tmp; \ + +#define R1_PRE(i) T1( W_PRE, i) +#define R1_MAIN(i) T1( W_MAIN, i) + +#if (!defined(_SHA256_UNROLL) || STEP_MAIN < 8) && (STEP_MAIN >= 4) +#define R2_MAIN(i) \ + R1_MAIN(i) \ + R1_MAIN(i + 1) \ + +#endif + + + +#if defined(_SHA256_UNROLL) && STEP_MAIN >= 8 + +#define T4( a,b,c,d,e,f,g,h, wx, i) \ + h += S1(e) + Ch(e,f,g) + K[(i)+(size_t)(j)] + wx(i); \ + tmp = h; \ + h += d; \ + d = tmp + S0(a) + Maj(a, b, c); \ + +#define R4( wx, i) \ + T4 ( a,b,c,d,e,f,g,h, wx, (i )); \ + T4 ( d,a,b,c,h,e,f,g, wx, (i+1)); \ + T4 ( c,d,a,b,g,h,e,f, wx, (i+2)); \ + T4 ( b,c,d,a,f,g,h,e, wx, (i+3)); \ + +#define R4_PRE(i) R4( W_PRE, i) +#define R4_MAIN(i) R4( W_MAIN, i) + + +#define T8( a,b,c,d,e,f,g,h, wx, i) \ + h += S1(e) + Ch(e,f,g) + K[(i)+(size_t)(j)] + wx(i); \ + d += h; \ + h += S0(a) + Maj(a, b, c); \ + +#define R8( wx, i) \ + T8 ( a,b,c,d,e,f,g,h, wx, i ); \ + T8 ( h,a,b,c,d,e,f,g, wx, i+1); \ + T8 ( g,h,a,b,c,d,e,f, wx, i+2); \ + T8 ( f,g,h,a,b,c,d,e, wx, i+3); \ + T8 ( e,f,g,h,a,b,c,d, wx, i+4); \ + T8 ( d,e,f,g,h,a,b,c, wx, i+5); \ + T8 ( c,d,e,f,g,h,a,b, wx, i+6); \ + T8 ( b,c,d,e,f,g,h,a, wx, i+7); \ + +#define R8_PRE(i) R8( W_PRE, i) +#define R8_MAIN(i) R8( W_MAIN, i) + +#endif + +void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); + +// static +extern MY_ALIGN(64) +const UInt32 SHA256_K_ARRAY[64]; + +MY_ALIGN(64) +const UInt32 SHA256_K_ARRAY[64] = { + 0x428a2f98, 0x71374491, 0xb5c0fbcf, 0xe9b5dba5, + 0x3956c25b, 0x59f111f1, 0x923f82a4, 0xab1c5ed5, + 0xd807aa98, 0x12835b01, 0x243185be, 0x550c7dc3, + 0x72be5d74, 0x80deb1fe, 0x9bdc06a7, 0xc19bf174, + 0xe49b69c1, 0xefbe4786, 0x0fc19dc6, 0x240ca1cc, + 0x2de92c6f, 0x4a7484aa, 0x5cb0a9dc, 0x76f988da, + 0x983e5152, 0xa831c66d, 0xb00327c8, 0xbf597fc7, + 0xc6e00bf3, 0xd5a79147, 0x06ca6351, 0x14292967, + 0x27b70a85, 0x2e1b2138, 0x4d2c6dfc, 0x53380d13, + 0x650a7354, 0x766a0abb, 0x81c2c92e, 0x92722c85, + 0xa2bfe8a1, 0xa81a664b, 0xc24b8b70, 0xc76c51a3, + 0xd192e819, 0xd6990624, 0xf40e3585, 0x106aa070, + 0x19a4c116, 0x1e376c08, 0x2748774c, 0x34b0bcb5, + 0x391c0cb3, 0x4ed8aa4a, 0x5b9cca4f, 0x682e6ff3, + 0x748f82ee, 0x78a5636f, 0x84c87814, 0x8cc70208, + 0x90befffa, 0xa4506ceb, 0xbef9a3f7, 0xc67178f2 +}; + +#define K SHA256_K_ARRAY + + +MY_NO_INLINE +void MY_FAST_CALL Sha256_UpdateBlocks(UInt32 state[8], const Byte *data, size_t numBlocks) +{ + UInt32 W + #ifdef _SHA256_BIG_W + [64]; + #else + [16]; + #endif + + unsigned j; + + UInt32 a,b,c,d,e,f,g,h; + + #if !defined(_SHA256_UNROLL) || (STEP_MAIN <= 4) || (STEP_PRE <= 4) + UInt32 tmp; + #endif + + a = state[0]; + b = state[1]; + c = state[2]; + d = state[3]; + e = state[4]; + f = state[5]; + g = state[6]; + h = state[7]; + + while (numBlocks) + { + + for (j = 0; j < 16; j += STEP_PRE) + { + #if STEP_PRE > 4 + + #if STEP_PRE < 8 + R4_PRE(0); + #else + R8_PRE(0); + #if STEP_PRE == 16 + R8_PRE(8); + #endif + #endif + + #else + + R1_PRE(0); + #if STEP_PRE >= 2 + R1_PRE(1); + #if STEP_PRE >= 4 + R1_PRE(2); + R1_PRE(3); + #endif + #endif + + #endif + } + + for (j = 16; j < 64; j += STEP_MAIN) + { + #if defined(_SHA256_UNROLL) && STEP_MAIN >= 8 + + #if STEP_MAIN < 8 + R4_MAIN(0); + #else + R8_MAIN(0); + #if STEP_MAIN == 16 + R8_MAIN(8); + #endif + #endif + + #else + + R1_MAIN(0); + #if STEP_MAIN >= 2 + R1_MAIN(1); + #if STEP_MAIN >= 4 + R2_MAIN(2); + #if STEP_MAIN >= 8 + R2_MAIN(4); + R2_MAIN(6); + #if STEP_MAIN >= 16 + R2_MAIN(8); + R2_MAIN(10); + R2_MAIN(12); + R2_MAIN(14); + #endif + #endif + #endif + #endif + #endif + } + + a += state[0]; state[0] = a; + b += state[1]; state[1] = b; + c += state[2]; state[2] = c; + d += state[3]; state[3] = d; + e += state[4]; state[4] = e; + f += state[5]; state[5] = f; + g += state[6]; state[6] = g; + h += state[7]; state[7] = h; + + data += 64; + numBlocks--; + } + + /* Wipe variables */ + /* memset(W, 0, sizeof(W)); */ +} + +#undef S0 +#undef S1 +#undef s0 +#undef s1 +#undef K + +#define Sha256_UpdateBlock(p) UPDATE_BLOCKS(p)(p->state, p->buffer, 1) + +void Sha256_Update(CSha256 *p, const Byte *data, size_t size) +{ + if (size == 0) + return; + + { + unsigned pos = (unsigned)p->count & 0x3F; + unsigned num; + + p->count += size; + + num = 64 - pos; + if (num > size) + { + memcpy(p->buffer + pos, data, size); + return; + } + + if (pos != 0) + { + size -= num; + memcpy(p->buffer + pos, data, num); + data += num; + Sha256_UpdateBlock(p); + } + } + { + size_t numBlocks = size >> 6; + UPDATE_BLOCKS(p)(p->state, data, numBlocks); + size &= 0x3F; + if (size == 0) + return; + data += (numBlocks << 6); + memcpy(p->buffer, data, size); + } +} + + +void Sha256_Final(CSha256 *p, Byte *digest) +{ + unsigned pos = (unsigned)p->count & 0x3F; + unsigned i; + + p->buffer[pos++] = 0x80; + + if (pos > (64 - 8)) + { + while (pos != 64) { p->buffer[pos++] = 0; } + // memset(&p->buf.buffer[pos], 0, 64 - pos); + Sha256_UpdateBlock(p); + pos = 0; + } + + /* + if (pos & 3) + { + p->buffer[pos] = 0; + p->buffer[pos + 1] = 0; + p->buffer[pos + 2] = 0; + pos += 3; + pos &= ~3; + } + { + for (; pos < 64 - 8; pos += 4) + *(UInt32 *)(&p->buffer[pos]) = 0; + } + */ + + memset(&p->buffer[pos], 0, (64 - 8) - pos); + + { + UInt64 numBits = (p->count << 3); + SetBe32(p->buffer + 64 - 8, (UInt32)(numBits >> 32)); + SetBe32(p->buffer + 64 - 4, (UInt32)(numBits)); + } + + Sha256_UpdateBlock(p); + + for (i = 0; i < 8; i += 2) + { + UInt32 v0 = p->state[i]; + UInt32 v1 = p->state[(size_t)i + 1]; + SetBe32(digest , v0); + SetBe32(digest + 4, v1); + digest += 8; + } + + Sha256_InitState(p); +} + + +void Sha256Prepare() +{ + #ifdef _SHA_SUPPORTED + SHA256_FUNC_UPDATE_BLOCKS f, f_hw; + f = Sha256_UpdateBlocks; + f_hw = NULL; + #ifdef MY_CPU_X86_OR_AMD64 + #ifndef USE_MY_MM + if (CPU_IsSupported_SHA() + && CPU_IsSupported_SSSE3() + // && CPU_IsSupported_SSE41() + ) + #endif + #else + if (CPU_IsSupported_SHA2()) + #endif + { + // printf("\n========== HW SHA256 ======== \n"); + f = f_hw = Sha256_UpdateBlocks_HW; + } + g_FUNC_UPDATE_BLOCKS = f; + g_FUNC_UPDATE_BLOCKS_HW = f_hw; + #endif +} diff --git a/deps/LZMA-SDK/C/Sha256.h b/deps/LZMA-SDK/C/Sha256.h index f52933986..aa38501e5 100644 --- a/deps/LZMA-SDK/C/Sha256.h +++ b/deps/LZMA-SDK/C/Sha256.h @@ -1,76 +1,76 @@ -/* Sha256.h -- SHA-256 Hash -2021-01-01 : Igor Pavlov : Public domain */ - -#ifndef __7Z_SHA256_H -#define __7Z_SHA256_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -#define SHA256_NUM_BLOCK_WORDS 16 -#define SHA256_NUM_DIGEST_WORDS 8 - -#define SHA256_BLOCK_SIZE (SHA256_NUM_BLOCK_WORDS * 4) -#define SHA256_DIGEST_SIZE (SHA256_NUM_DIGEST_WORDS * 4) - -typedef void (MY_FAST_CALL *SHA256_FUNC_UPDATE_BLOCKS)(UInt32 state[8], const Byte *data, size_t numBlocks); - -/* - if (the system supports different SHA256 code implementations) - { - (CSha256::func_UpdateBlocks) will be used - (CSha256::func_UpdateBlocks) can be set by - Sha256_Init() - to default (fastest) - Sha256_SetFunction() - to any algo - } - else - { - (CSha256::func_UpdateBlocks) is ignored. - } -*/ - -typedef struct -{ - SHA256_FUNC_UPDATE_BLOCKS func_UpdateBlocks; - UInt64 count; - UInt64 __pad_2[2]; - UInt32 state[SHA256_NUM_DIGEST_WORDS]; - - Byte buffer[SHA256_BLOCK_SIZE]; -} CSha256; - - -#define SHA256_ALGO_DEFAULT 0 -#define SHA256_ALGO_SW 1 -#define SHA256_ALGO_HW 2 - -/* -Sha256_SetFunction() -return: - 0 - (algo) value is not supported, and func_UpdateBlocks was not changed - 1 - func_UpdateBlocks was set according (algo) value. -*/ - -BoolInt Sha256_SetFunction(CSha256 *p, unsigned algo); - -void Sha256_InitState(CSha256 *p); -void Sha256_Init(CSha256 *p); -void Sha256_Update(CSha256 *p, const Byte *data, size_t size); -void Sha256_Final(CSha256 *p, Byte *digest); - - - - -// void MY_FAST_CALL Sha256_UpdateBlocks(UInt32 state[8], const Byte *data, size_t numBlocks); - -/* -call Sha256Prepare() once at program start. -It prepares all supported implementations, and detects the fastest implementation. -*/ - -void Sha256Prepare(void); - -EXTERN_C_END - -#endif +/* Sha256.h -- SHA-256 Hash +2021-01-01 : Igor Pavlov : Public domain */ + +#ifndef __7Z_SHA256_H +#define __7Z_SHA256_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define SHA256_NUM_BLOCK_WORDS 16 +#define SHA256_NUM_DIGEST_WORDS 8 + +#define SHA256_BLOCK_SIZE (SHA256_NUM_BLOCK_WORDS * 4) +#define SHA256_DIGEST_SIZE (SHA256_NUM_DIGEST_WORDS * 4) + +typedef void (MY_FAST_CALL *SHA256_FUNC_UPDATE_BLOCKS)(UInt32 state[8], const Byte *data, size_t numBlocks); + +/* + if (the system supports different SHA256 code implementations) + { + (CSha256::func_UpdateBlocks) will be used + (CSha256::func_UpdateBlocks) can be set by + Sha256_Init() - to default (fastest) + Sha256_SetFunction() - to any algo + } + else + { + (CSha256::func_UpdateBlocks) is ignored. + } +*/ + +typedef struct +{ + SHA256_FUNC_UPDATE_BLOCKS func_UpdateBlocks; + UInt64 count; + UInt64 __pad_2[2]; + UInt32 state[SHA256_NUM_DIGEST_WORDS]; + + Byte buffer[SHA256_BLOCK_SIZE]; +} CSha256; + + +#define SHA256_ALGO_DEFAULT 0 +#define SHA256_ALGO_SW 1 +#define SHA256_ALGO_HW 2 + +/* +Sha256_SetFunction() +return: + 0 - (algo) value is not supported, and func_UpdateBlocks was not changed + 1 - func_UpdateBlocks was set according (algo) value. +*/ + +BoolInt Sha256_SetFunction(CSha256 *p, unsigned algo); + +void Sha256_InitState(CSha256 *p); +void Sha256_Init(CSha256 *p); +void Sha256_Update(CSha256 *p, const Byte *data, size_t size); +void Sha256_Final(CSha256 *p, Byte *digest); + + + + +// void MY_FAST_CALL Sha256_UpdateBlocks(UInt32 state[8], const Byte *data, size_t numBlocks); + +/* +call Sha256Prepare() once at program start. +It prepares all supported implementations, and detects the fastest implementation. +*/ + +void Sha256Prepare(void); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/Sha256Opt.c b/deps/LZMA-SDK/C/Sha256Opt.c index cc8c53e1b..decc1382c 100644 --- a/deps/LZMA-SDK/C/Sha256Opt.c +++ b/deps/LZMA-SDK/C/Sha256Opt.c @@ -1,373 +1,373 @@ -/* Sha256Opt.c -- SHA-256 optimized code for SHA-256 hardware instructions -2021-04-01 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#if defined(_MSC_VER) -#if (_MSC_VER < 1900) && (_MSC_VER >= 1200) -// #define USE_MY_MM -#endif -#endif - -#include "CpuArch.h" - -#ifdef MY_CPU_X86_OR_AMD64 - #if defined(__clang__) - #if (__clang_major__ >= 8) // fix that check - #define USE_HW_SHA - #ifndef __SHA__ - #define ATTRIB_SHA __attribute__((__target__("sha,ssse3"))) - #if defined(_MSC_VER) - // SSSE3: for clang-cl: - #include - #define __SHA__ - #endif - #endif - - #endif - #elif defined(__GNUC__) - #if (__GNUC__ >= 8) // fix that check - #define USE_HW_SHA - #ifndef __SHA__ - #define ATTRIB_SHA __attribute__((__target__("sha,ssse3"))) - // #pragma GCC target("sha,ssse3") - #endif - #endif - #elif defined(__INTEL_COMPILER) - #if (__INTEL_COMPILER >= 1800) // fix that check - #define USE_HW_SHA - #endif - #elif defined(_MSC_VER) - #ifdef USE_MY_MM - #define USE_VER_MIN 1300 - #else - #define USE_VER_MIN 1910 - #endif - #if _MSC_VER >= USE_VER_MIN - #define USE_HW_SHA - #endif - #endif -// #endif // MY_CPU_X86_OR_AMD64 - -#ifdef USE_HW_SHA - -// #pragma message("Sha256 HW") -// #include - -#if !defined(_MSC_VER) || (_MSC_VER >= 1900) -#include -#else -#include - -#if defined(_MSC_VER) && (_MSC_VER >= 1600) -// #include -#endif - -#ifdef USE_MY_MM -#include "My_mm.h" -#endif - -#endif - -/* -SHA256 uses: -SSE2: - _mm_loadu_si128 - _mm_storeu_si128 - _mm_set_epi32 - _mm_add_epi32 - _mm_shuffle_epi32 / pshufd - - - -SSSE3: - _mm_shuffle_epi8 / pshufb - _mm_alignr_epi8 -SHA: - _mm_sha256* -*/ - -// K array must be aligned for 16-bytes at least. -// The compiler can look align attribute and selects -// movdqu - for code without align attribute -// movdqa - for code with align attribute -extern -MY_ALIGN(64) -const UInt32 SHA256_K_ARRAY[64]; - -#define K SHA256_K_ARRAY - - -#define ADD_EPI32(dest, src) dest = _mm_add_epi32(dest, src); -#define SHA256_MSG1(dest, src) dest = _mm_sha256msg1_epu32(dest, src); -#define SHA25G_MSG2(dest, src) dest = _mm_sha256msg2_epu32(dest, src); - - -#define LOAD_SHUFFLE(m, k) \ - m = _mm_loadu_si128((const __m128i *)(const void *)(data + (k) * 16)); \ - m = _mm_shuffle_epi8(m, mask); \ - -#define SM1(g0, g1, g2, g3) \ - SHA256_MSG1(g3, g0); \ - -#define SM2(g0, g1, g2, g3) \ - tmp = _mm_alignr_epi8(g1, g0, 4); \ - ADD_EPI32(g2, tmp); \ - SHA25G_MSG2(g2, g1); \ - -// #define LS0(k, g0, g1, g2, g3) LOAD_SHUFFLE(g0, k) -// #define LS1(k, g0, g1, g2, g3) LOAD_SHUFFLE(g1, k+1) - - -#define NNN(g0, g1, g2, g3) - - -#define RND2(t0, t1) \ - t0 = _mm_sha256rnds2_epu32(t0, t1, msg); - -#define RND2_0(m, k) \ - msg = _mm_add_epi32(m, *(const __m128i *) (const void *) &K[(k) * 4]); \ - RND2(state0, state1); \ - msg = _mm_shuffle_epi32(msg, 0x0E); \ - - -#define RND2_1 \ - RND2(state1, state0); \ - - -// We use scheme with 3 rounds ahead for SHA256_MSG1 / 2 rounds ahead for SHA256_MSG2 - -#define R4(k, g0, g1, g2, g3, OP0, OP1) \ - RND2_0(g0, k); \ - OP0(g0, g1, g2, g3); \ - RND2_1; \ - OP1(g0, g1, g2, g3); \ - -#define R16(k, OP0, OP1, OP2, OP3, OP4, OP5, OP6, OP7) \ - R4 ( (k)*4+0, m0, m1, m2, m3, OP0, OP1 ) \ - R4 ( (k)*4+1, m1, m2, m3, m0, OP2, OP3 ) \ - R4 ( (k)*4+2, m2, m3, m0, m1, OP4, OP5 ) \ - R4 ( (k)*4+3, m3, m0, m1, m2, OP6, OP7 ) \ - -#define PREPARE_STATE \ - tmp = _mm_shuffle_epi32(state0, 0x1B); /* abcd */ \ - state0 = _mm_shuffle_epi32(state1, 0x1B); /* efgh */ \ - state1 = state0; \ - state0 = _mm_unpacklo_epi64(state0, tmp); /* cdgh */ \ - state1 = _mm_unpackhi_epi64(state1, tmp); /* abef */ \ - - -void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); -#ifdef ATTRIB_SHA -ATTRIB_SHA -#endif -void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks) -{ - const __m128i mask = _mm_set_epi32(0x0c0d0e0f, 0x08090a0b, 0x04050607, 0x00010203); - __m128i tmp; - __m128i state0, state1; - - if (numBlocks == 0) - return; - - state0 = _mm_loadu_si128((const __m128i *) (const void *) &state[0]); - state1 = _mm_loadu_si128((const __m128i *) (const void *) &state[4]); - - PREPARE_STATE - - do - { - __m128i state0_save, state1_save; - __m128i m0, m1, m2, m3; - __m128i msg; - // #define msg tmp - - state0_save = state0; - state1_save = state1; - - LOAD_SHUFFLE (m0, 0) - LOAD_SHUFFLE (m1, 1) - LOAD_SHUFFLE (m2, 2) - LOAD_SHUFFLE (m3, 3) - - - - R16 ( 0, NNN, NNN, SM1, NNN, SM1, SM2, SM1, SM2 ); - R16 ( 1, SM1, SM2, SM1, SM2, SM1, SM2, SM1, SM2 ); - R16 ( 2, SM1, SM2, SM1, SM2, SM1, SM2, SM1, SM2 ); - R16 ( 3, SM1, SM2, NNN, SM2, NNN, NNN, NNN, NNN ); - - ADD_EPI32(state0, state0_save); - ADD_EPI32(state1, state1_save); - - data += 64; - } - while (--numBlocks); - - PREPARE_STATE - - _mm_storeu_si128((__m128i *) (void *) &state[0], state0); - _mm_storeu_si128((__m128i *) (void *) &state[4], state1); -} - -#endif // USE_HW_SHA - -#elif defined(MY_CPU_ARM_OR_ARM64) - - #if defined(__clang__) - #if (__clang_major__ >= 8) // fix that check - #define USE_HW_SHA - #endif - #elif defined(__GNUC__) - #if (__GNUC__ >= 6) // fix that check - #define USE_HW_SHA - #endif - #elif defined(_MSC_VER) - #if _MSC_VER >= 1910 - #define USE_HW_SHA - #endif - #endif - -#ifdef USE_HW_SHA - -// #pragma message("=== Sha256 HW === ") - -#if defined(__clang__) || defined(__GNUC__) - #ifdef MY_CPU_ARM64 - #define ATTRIB_SHA __attribute__((__target__("+crypto"))) - #else - #define ATTRIB_SHA __attribute__((__target__("fpu=crypto-neon-fp-armv8"))) - #endif -#else - // _MSC_VER - // for arm32 - #define _ARM_USE_NEW_NEON_INTRINSICS -#endif - -#if defined(_MSC_VER) && defined(MY_CPU_ARM64) -#include -#else -#include -#endif - -typedef uint32x4_t v128; -// typedef __n128 v128; // MSVC - -#ifdef MY_CPU_BE - #define MY_rev32_for_LE(x) -#else - #define MY_rev32_for_LE(x) x = vreinterpretq_u32_u8(vrev32q_u8(vreinterpretq_u8_u32(x))) -#endif - -#define LOAD_128(_p) (*(const v128 *)(const void *)(_p)) -#define STORE_128(_p, _v) *(v128 *)(void *)(_p) = (_v) - -#define LOAD_SHUFFLE(m, k) \ - m = LOAD_128((data + (k) * 16)); \ - MY_rev32_for_LE(m); \ - -// K array must be aligned for 16-bytes at least. -extern -MY_ALIGN(64) -const UInt32 SHA256_K_ARRAY[64]; - -#define K SHA256_K_ARRAY - - -#define SHA256_SU0(dest, src) dest = vsha256su0q_u32(dest, src); -#define SHA25G_SU1(dest, src2, src3) dest = vsha256su1q_u32(dest, src2, src3); - -#define SM1(g0, g1, g2, g3) SHA256_SU0(g3, g0) -#define SM2(g0, g1, g2, g3) SHA25G_SU1(g2, g0, g1) -#define NNN(g0, g1, g2, g3) - - -#define R4(k, g0, g1, g2, g3, OP0, OP1) \ - msg = vaddq_u32(g0, *(const v128 *) (const void *) &K[(k) * 4]); \ - tmp = state0; \ - state0 = vsha256hq_u32( state0, state1, msg ); \ - state1 = vsha256h2q_u32( state1, tmp, msg ); \ - OP0(g0, g1, g2, g3); \ - OP1(g0, g1, g2, g3); \ - - -#define R16(k, OP0, OP1, OP2, OP3, OP4, OP5, OP6, OP7) \ - R4 ( (k)*4+0, m0, m1, m2, m3, OP0, OP1 ) \ - R4 ( (k)*4+1, m1, m2, m3, m0, OP2, OP3 ) \ - R4 ( (k)*4+2, m2, m3, m0, m1, OP4, OP5 ) \ - R4 ( (k)*4+3, m3, m0, m1, m2, OP6, OP7 ) \ - - -void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); -#ifdef ATTRIB_SHA -ATTRIB_SHA -#endif -void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks) -{ - v128 state0, state1; - - if (numBlocks == 0) - return; - - state0 = LOAD_128(&state[0]); - state1 = LOAD_128(&state[4]); - - do - { - v128 state0_save, state1_save; - v128 m0, m1, m2, m3; - v128 msg, tmp; - - state0_save = state0; - state1_save = state1; - - LOAD_SHUFFLE (m0, 0) - LOAD_SHUFFLE (m1, 1) - LOAD_SHUFFLE (m2, 2) - LOAD_SHUFFLE (m3, 3) - - R16 ( 0, NNN, NNN, SM1, NNN, SM1, SM2, SM1, SM2 ); - R16 ( 1, SM1, SM2, SM1, SM2, SM1, SM2, SM1, SM2 ); - R16 ( 2, SM1, SM2, SM1, SM2, SM1, SM2, SM1, SM2 ); - R16 ( 3, SM1, SM2, NNN, SM2, NNN, NNN, NNN, NNN ); - - state0 = vaddq_u32(state0, state0_save); - state1 = vaddq_u32(state1, state1_save); - - data += 64; - } - while (--numBlocks); - - STORE_128(&state[0], state0); - STORE_128(&state[4], state1); -} - -#endif // USE_HW_SHA - -#endif // MY_CPU_ARM_OR_ARM64 - - -#ifndef USE_HW_SHA - -// #error Stop_Compiling_UNSUPPORTED_SHA -// #include - -// #include "Sha256.h" -void MY_FAST_CALL Sha256_UpdateBlocks(UInt32 state[8], const Byte *data, size_t numBlocks); - -#pragma message("Sha256 HW-SW stub was used") - -void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); -void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks) -{ - Sha256_UpdateBlocks(state, data, numBlocks); - /* - UNUSED_VAR(state); - UNUSED_VAR(data); - UNUSED_VAR(numBlocks); - exit(1); - return; - */ -} - -#endif +/* Sha256Opt.c -- SHA-256 optimized code for SHA-256 hardware instructions +2021-04-01 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#if defined(_MSC_VER) +#if (_MSC_VER < 1900) && (_MSC_VER >= 1200) +// #define USE_MY_MM +#endif +#endif + +#include "CpuArch.h" + +#ifdef MY_CPU_X86_OR_AMD64 + #if defined(__clang__) + #if (__clang_major__ >= 8) // fix that check + #define USE_HW_SHA + #ifndef __SHA__ + #define ATTRIB_SHA __attribute__((__target__("sha,ssse3"))) + #if defined(_MSC_VER) + // SSSE3: for clang-cl: + #include + #define __SHA__ + #endif + #endif + + #endif + #elif defined(__GNUC__) + #if (__GNUC__ >= 8) // fix that check + #define USE_HW_SHA + #ifndef __SHA__ + #define ATTRIB_SHA __attribute__((__target__("sha,ssse3"))) + // #pragma GCC target("sha,ssse3") + #endif + #endif + #elif defined(__INTEL_COMPILER) + #if (__INTEL_COMPILER >= 1800) // fix that check + #define USE_HW_SHA + #endif + #elif defined(_MSC_VER) + #ifdef USE_MY_MM + #define USE_VER_MIN 1300 + #else + #define USE_VER_MIN 1910 + #endif + #if _MSC_VER >= USE_VER_MIN + #define USE_HW_SHA + #endif + #endif +// #endif // MY_CPU_X86_OR_AMD64 + +#ifdef USE_HW_SHA + +// #pragma message("Sha256 HW") +// #include + +#if !defined(_MSC_VER) || (_MSC_VER >= 1900) +#include +#else +#include + +#if defined(_MSC_VER) && (_MSC_VER >= 1600) +// #include +#endif + +#ifdef USE_MY_MM +#include "My_mm.h" +#endif + +#endif + +/* +SHA256 uses: +SSE2: + _mm_loadu_si128 + _mm_storeu_si128 + _mm_set_epi32 + _mm_add_epi32 + _mm_shuffle_epi32 / pshufd + + + +SSSE3: + _mm_shuffle_epi8 / pshufb + _mm_alignr_epi8 +SHA: + _mm_sha256* +*/ + +// K array must be aligned for 16-bytes at least. +// The compiler can look align attribute and selects +// movdqu - for code without align attribute +// movdqa - for code with align attribute +extern +MY_ALIGN(64) +const UInt32 SHA256_K_ARRAY[64]; + +#define K SHA256_K_ARRAY + + +#define ADD_EPI32(dest, src) dest = _mm_add_epi32(dest, src); +#define SHA256_MSG1(dest, src) dest = _mm_sha256msg1_epu32(dest, src); +#define SHA25G_MSG2(dest, src) dest = _mm_sha256msg2_epu32(dest, src); + + +#define LOAD_SHUFFLE(m, k) \ + m = _mm_loadu_si128((const __m128i *)(const void *)(data + (k) * 16)); \ + m = _mm_shuffle_epi8(m, mask); \ + +#define SM1(g0, g1, g2, g3) \ + SHA256_MSG1(g3, g0); \ + +#define SM2(g0, g1, g2, g3) \ + tmp = _mm_alignr_epi8(g1, g0, 4); \ + ADD_EPI32(g2, tmp); \ + SHA25G_MSG2(g2, g1); \ + +// #define LS0(k, g0, g1, g2, g3) LOAD_SHUFFLE(g0, k) +// #define LS1(k, g0, g1, g2, g3) LOAD_SHUFFLE(g1, k+1) + + +#define NNN(g0, g1, g2, g3) + + +#define RND2(t0, t1) \ + t0 = _mm_sha256rnds2_epu32(t0, t1, msg); + +#define RND2_0(m, k) \ + msg = _mm_add_epi32(m, *(const __m128i *) (const void *) &K[(k) * 4]); \ + RND2(state0, state1); \ + msg = _mm_shuffle_epi32(msg, 0x0E); \ + + +#define RND2_1 \ + RND2(state1, state0); \ + + +// We use scheme with 3 rounds ahead for SHA256_MSG1 / 2 rounds ahead for SHA256_MSG2 + +#define R4(k, g0, g1, g2, g3, OP0, OP1) \ + RND2_0(g0, k); \ + OP0(g0, g1, g2, g3); \ + RND2_1; \ + OP1(g0, g1, g2, g3); \ + +#define R16(k, OP0, OP1, OP2, OP3, OP4, OP5, OP6, OP7) \ + R4 ( (k)*4+0, m0, m1, m2, m3, OP0, OP1 ) \ + R4 ( (k)*4+1, m1, m2, m3, m0, OP2, OP3 ) \ + R4 ( (k)*4+2, m2, m3, m0, m1, OP4, OP5 ) \ + R4 ( (k)*4+3, m3, m0, m1, m2, OP6, OP7 ) \ + +#define PREPARE_STATE \ + tmp = _mm_shuffle_epi32(state0, 0x1B); /* abcd */ \ + state0 = _mm_shuffle_epi32(state1, 0x1B); /* efgh */ \ + state1 = state0; \ + state0 = _mm_unpacklo_epi64(state0, tmp); /* cdgh */ \ + state1 = _mm_unpackhi_epi64(state1, tmp); /* abef */ \ + + +void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); +#ifdef ATTRIB_SHA +ATTRIB_SHA +#endif +void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks) +{ + const __m128i mask = _mm_set_epi32(0x0c0d0e0f, 0x08090a0b, 0x04050607, 0x00010203); + __m128i tmp; + __m128i state0, state1; + + if (numBlocks == 0) + return; + + state0 = _mm_loadu_si128((const __m128i *) (const void *) &state[0]); + state1 = _mm_loadu_si128((const __m128i *) (const void *) &state[4]); + + PREPARE_STATE + + do + { + __m128i state0_save, state1_save; + __m128i m0, m1, m2, m3; + __m128i msg; + // #define msg tmp + + state0_save = state0; + state1_save = state1; + + LOAD_SHUFFLE (m0, 0) + LOAD_SHUFFLE (m1, 1) + LOAD_SHUFFLE (m2, 2) + LOAD_SHUFFLE (m3, 3) + + + + R16 ( 0, NNN, NNN, SM1, NNN, SM1, SM2, SM1, SM2 ); + R16 ( 1, SM1, SM2, SM1, SM2, SM1, SM2, SM1, SM2 ); + R16 ( 2, SM1, SM2, SM1, SM2, SM1, SM2, SM1, SM2 ); + R16 ( 3, SM1, SM2, NNN, SM2, NNN, NNN, NNN, NNN ); + + ADD_EPI32(state0, state0_save); + ADD_EPI32(state1, state1_save); + + data += 64; + } + while (--numBlocks); + + PREPARE_STATE + + _mm_storeu_si128((__m128i *) (void *) &state[0], state0); + _mm_storeu_si128((__m128i *) (void *) &state[4], state1); +} + +#endif // USE_HW_SHA + +#elif defined(MY_CPU_ARM_OR_ARM64) + + #if defined(__clang__) + #if (__clang_major__ >= 8) // fix that check + #define USE_HW_SHA + #endif + #elif defined(__GNUC__) + #if (__GNUC__ >= 6) // fix that check + #define USE_HW_SHA + #endif + #elif defined(_MSC_VER) + #if _MSC_VER >= 1910 + #define USE_HW_SHA + #endif + #endif + +#ifdef USE_HW_SHA + +// #pragma message("=== Sha256 HW === ") + +#if defined(__clang__) || defined(__GNUC__) + #ifdef MY_CPU_ARM64 + #define ATTRIB_SHA __attribute__((__target__("+crypto"))) + #else + #define ATTRIB_SHA __attribute__((__target__("fpu=crypto-neon-fp-armv8"))) + #endif +#else + // _MSC_VER + // for arm32 + #define _ARM_USE_NEW_NEON_INTRINSICS +#endif + +#if defined(_MSC_VER) && defined(MY_CPU_ARM64) +#include +#else +#include +#endif + +typedef uint32x4_t v128; +// typedef __n128 v128; // MSVC + +#ifdef MY_CPU_BE + #define MY_rev32_for_LE(x) +#else + #define MY_rev32_for_LE(x) x = vreinterpretq_u32_u8(vrev32q_u8(vreinterpretq_u8_u32(x))) +#endif + +#define LOAD_128(_p) (*(const v128 *)(const void *)(_p)) +#define STORE_128(_p, _v) *(v128 *)(void *)(_p) = (_v) + +#define LOAD_SHUFFLE(m, k) \ + m = LOAD_128((data + (k) * 16)); \ + MY_rev32_for_LE(m); \ + +// K array must be aligned for 16-bytes at least. +extern +MY_ALIGN(64) +const UInt32 SHA256_K_ARRAY[64]; + +#define K SHA256_K_ARRAY + + +#define SHA256_SU0(dest, src) dest = vsha256su0q_u32(dest, src); +#define SHA25G_SU1(dest, src2, src3) dest = vsha256su1q_u32(dest, src2, src3); + +#define SM1(g0, g1, g2, g3) SHA256_SU0(g3, g0) +#define SM2(g0, g1, g2, g3) SHA25G_SU1(g2, g0, g1) +#define NNN(g0, g1, g2, g3) + + +#define R4(k, g0, g1, g2, g3, OP0, OP1) \ + msg = vaddq_u32(g0, *(const v128 *) (const void *) &K[(k) * 4]); \ + tmp = state0; \ + state0 = vsha256hq_u32( state0, state1, msg ); \ + state1 = vsha256h2q_u32( state1, tmp, msg ); \ + OP0(g0, g1, g2, g3); \ + OP1(g0, g1, g2, g3); \ + + +#define R16(k, OP0, OP1, OP2, OP3, OP4, OP5, OP6, OP7) \ + R4 ( (k)*4+0, m0, m1, m2, m3, OP0, OP1 ) \ + R4 ( (k)*4+1, m1, m2, m3, m0, OP2, OP3 ) \ + R4 ( (k)*4+2, m2, m3, m0, m1, OP4, OP5 ) \ + R4 ( (k)*4+3, m3, m0, m1, m2, OP6, OP7 ) \ + + +void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); +#ifdef ATTRIB_SHA +ATTRIB_SHA +#endif +void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks) +{ + v128 state0, state1; + + if (numBlocks == 0) + return; + + state0 = LOAD_128(&state[0]); + state1 = LOAD_128(&state[4]); + + do + { + v128 state0_save, state1_save; + v128 m0, m1, m2, m3; + v128 msg, tmp; + + state0_save = state0; + state1_save = state1; + + LOAD_SHUFFLE (m0, 0) + LOAD_SHUFFLE (m1, 1) + LOAD_SHUFFLE (m2, 2) + LOAD_SHUFFLE (m3, 3) + + R16 ( 0, NNN, NNN, SM1, NNN, SM1, SM2, SM1, SM2 ); + R16 ( 1, SM1, SM2, SM1, SM2, SM1, SM2, SM1, SM2 ); + R16 ( 2, SM1, SM2, SM1, SM2, SM1, SM2, SM1, SM2 ); + R16 ( 3, SM1, SM2, NNN, SM2, NNN, NNN, NNN, NNN ); + + state0 = vaddq_u32(state0, state0_save); + state1 = vaddq_u32(state1, state1_save); + + data += 64; + } + while (--numBlocks); + + STORE_128(&state[0], state0); + STORE_128(&state[4], state1); +} + +#endif // USE_HW_SHA + +#endif // MY_CPU_ARM_OR_ARM64 + + +#ifndef USE_HW_SHA + +// #error Stop_Compiling_UNSUPPORTED_SHA +// #include + +// #include "Sha256.h" +void MY_FAST_CALL Sha256_UpdateBlocks(UInt32 state[8], const Byte *data, size_t numBlocks); + +#pragma message("Sha256 HW-SW stub was used") + +void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); +void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks) +{ + Sha256_UpdateBlocks(state, data, numBlocks); + /* + UNUSED_VAR(state); + UNUSED_VAR(data); + UNUSED_VAR(numBlocks); + exit(1); + return; + */ +} + +#endif diff --git a/deps/LZMA-SDK/C/Sort.c b/deps/LZMA-SDK/C/Sort.c index 73dcbf059..e1097e380 100644 --- a/deps/LZMA-SDK/C/Sort.c +++ b/deps/LZMA-SDK/C/Sort.c @@ -1,141 +1,141 @@ -/* Sort.c -- Sort functions -2014-04-05 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "Sort.h" - -#define HeapSortDown(p, k, size, temp) \ - { for (;;) { \ - size_t s = (k << 1); \ - if (s > size) break; \ - if (s < size && p[s + 1] > p[s]) s++; \ - if (temp >= p[s]) break; \ - p[k] = p[s]; k = s; \ - } p[k] = temp; } - -void HeapSort(UInt32 *p, size_t size) -{ - if (size <= 1) - return; - p--; - { - size_t i = size / 2; - do - { - UInt32 temp = p[i]; - size_t k = i; - HeapSortDown(p, k, size, temp) - } - while (--i != 0); - } - /* - do - { - size_t k = 1; - UInt32 temp = p[size]; - p[size--] = p[1]; - HeapSortDown(p, k, size, temp) - } - while (size > 1); - */ - while (size > 3) - { - UInt32 temp = p[size]; - size_t k = (p[3] > p[2]) ? 3 : 2; - p[size--] = p[1]; - p[1] = p[k]; - HeapSortDown(p, k, size, temp) - } - { - UInt32 temp = p[size]; - p[size] = p[1]; - if (size > 2 && p[2] < temp) - { - p[1] = p[2]; - p[2] = temp; - } - else - p[1] = temp; - } -} - -void HeapSort64(UInt64 *p, size_t size) -{ - if (size <= 1) - return; - p--; - { - size_t i = size / 2; - do - { - UInt64 temp = p[i]; - size_t k = i; - HeapSortDown(p, k, size, temp) - } - while (--i != 0); - } - /* - do - { - size_t k = 1; - UInt64 temp = p[size]; - p[size--] = p[1]; - HeapSortDown(p, k, size, temp) - } - while (size > 1); - */ - while (size > 3) - { - UInt64 temp = p[size]; - size_t k = (p[3] > p[2]) ? 3 : 2; - p[size--] = p[1]; - p[1] = p[k]; - HeapSortDown(p, k, size, temp) - } - { - UInt64 temp = p[size]; - p[size] = p[1]; - if (size > 2 && p[2] < temp) - { - p[1] = p[2]; - p[2] = temp; - } - else - p[1] = temp; - } -} - -/* -#define HeapSortRefDown(p, vals, n, size, temp) \ - { size_t k = n; UInt32 val = vals[temp]; for (;;) { \ - size_t s = (k << 1); \ - if (s > size) break; \ - if (s < size && vals[p[s + 1]] > vals[p[s]]) s++; \ - if (val >= vals[p[s]]) break; \ - p[k] = p[s]; k = s; \ - } p[k] = temp; } - -void HeapSortRef(UInt32 *p, UInt32 *vals, size_t size) -{ - if (size <= 1) - return; - p--; - { - size_t i = size / 2; - do - { - UInt32 temp = p[i]; - HeapSortRefDown(p, vals, i, size, temp); - } - while (--i != 0); - } - do - { - UInt32 temp = p[size]; - p[size--] = p[1]; - HeapSortRefDown(p, vals, 1, size, temp); - } - while (size > 1); -} -*/ +/* Sort.c -- Sort functions +2014-04-05 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "Sort.h" + +#define HeapSortDown(p, k, size, temp) \ + { for (;;) { \ + size_t s = (k << 1); \ + if (s > size) break; \ + if (s < size && p[s + 1] > p[s]) s++; \ + if (temp >= p[s]) break; \ + p[k] = p[s]; k = s; \ + } p[k] = temp; } + +void HeapSort(UInt32 *p, size_t size) +{ + if (size <= 1) + return; + p--; + { + size_t i = size / 2; + do + { + UInt32 temp = p[i]; + size_t k = i; + HeapSortDown(p, k, size, temp) + } + while (--i != 0); + } + /* + do + { + size_t k = 1; + UInt32 temp = p[size]; + p[size--] = p[1]; + HeapSortDown(p, k, size, temp) + } + while (size > 1); + */ + while (size > 3) + { + UInt32 temp = p[size]; + size_t k = (p[3] > p[2]) ? 3 : 2; + p[size--] = p[1]; + p[1] = p[k]; + HeapSortDown(p, k, size, temp) + } + { + UInt32 temp = p[size]; + p[size] = p[1]; + if (size > 2 && p[2] < temp) + { + p[1] = p[2]; + p[2] = temp; + } + else + p[1] = temp; + } +} + +void HeapSort64(UInt64 *p, size_t size) +{ + if (size <= 1) + return; + p--; + { + size_t i = size / 2; + do + { + UInt64 temp = p[i]; + size_t k = i; + HeapSortDown(p, k, size, temp) + } + while (--i != 0); + } + /* + do + { + size_t k = 1; + UInt64 temp = p[size]; + p[size--] = p[1]; + HeapSortDown(p, k, size, temp) + } + while (size > 1); + */ + while (size > 3) + { + UInt64 temp = p[size]; + size_t k = (p[3] > p[2]) ? 3 : 2; + p[size--] = p[1]; + p[1] = p[k]; + HeapSortDown(p, k, size, temp) + } + { + UInt64 temp = p[size]; + p[size] = p[1]; + if (size > 2 && p[2] < temp) + { + p[1] = p[2]; + p[2] = temp; + } + else + p[1] = temp; + } +} + +/* +#define HeapSortRefDown(p, vals, n, size, temp) \ + { size_t k = n; UInt32 val = vals[temp]; for (;;) { \ + size_t s = (k << 1); \ + if (s > size) break; \ + if (s < size && vals[p[s + 1]] > vals[p[s]]) s++; \ + if (val >= vals[p[s]]) break; \ + p[k] = p[s]; k = s; \ + } p[k] = temp; } + +void HeapSortRef(UInt32 *p, UInt32 *vals, size_t size) +{ + if (size <= 1) + return; + p--; + { + size_t i = size / 2; + do + { + UInt32 temp = p[i]; + HeapSortRefDown(p, vals, i, size, temp); + } + while (--i != 0); + } + do + { + UInt32 temp = p[size]; + p[size--] = p[1]; + HeapSortRefDown(p, vals, 1, size, temp); + } + while (size > 1); +} +*/ diff --git a/deps/LZMA-SDK/C/Sort.h b/deps/LZMA-SDK/C/Sort.h index 7209d7824..2e2963a23 100644 --- a/deps/LZMA-SDK/C/Sort.h +++ b/deps/LZMA-SDK/C/Sort.h @@ -1,18 +1,18 @@ -/* Sort.h -- Sort functions -2014-04-05 : Igor Pavlov : Public domain */ - -#ifndef __7Z_SORT_H -#define __7Z_SORT_H - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -void HeapSort(UInt32 *p, size_t size); -void HeapSort64(UInt64 *p, size_t size); - -/* void HeapSortRef(UInt32 *p, UInt32 *vals, size_t size); */ - -EXTERN_C_END - -#endif +/* Sort.h -- Sort functions +2014-04-05 : Igor Pavlov : Public domain */ + +#ifndef __7Z_SORT_H +#define __7Z_SORT_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +void HeapSort(UInt32 *p, size_t size); +void HeapSort64(UInt64 *p, size_t size); + +/* void HeapSortRef(UInt32 *p, UInt32 *vals, size_t size); */ + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/Threads.c b/deps/LZMA-SDK/C/Threads.c index 6eb45b08a..58eb90ffa 100644 --- a/deps/LZMA-SDK/C/Threads.c +++ b/deps/LZMA-SDK/C/Threads.c @@ -1,540 +1,540 @@ -/* Threads.c -- multithreading library -2021-12-21 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#ifdef _WIN32 - -#ifndef USE_THREADS_CreateThread -#include -#endif - -#include "Threads.h" - -static WRes GetError() -{ - DWORD res = GetLastError(); - return res ? (WRes)res : 1; -} - -static WRes HandleToWRes(HANDLE h) { return (h != NULL) ? 0 : GetError(); } -static WRes BOOLToWRes(BOOL v) { return v ? 0 : GetError(); } - -WRes HandlePtr_Close(HANDLE *p) -{ - if (*p != NULL) - { - if (!CloseHandle(*p)) - return GetError(); - *p = NULL; - } - return 0; -} - -WRes Handle_WaitObject(HANDLE h) -{ - DWORD dw = WaitForSingleObject(h, INFINITE); - /* - (dw) result: - WAIT_OBJECT_0 // 0 - WAIT_ABANDONED // 0x00000080 : is not compatible with Win32 Error space - WAIT_TIMEOUT // 0x00000102 : is compatible with Win32 Error space - WAIT_FAILED // 0xFFFFFFFF - */ - if (dw == WAIT_FAILED) - { - dw = GetLastError(); - if (dw == 0) - return WAIT_FAILED; - } - return (WRes)dw; -} - -#define Thread_Wait(p) Handle_WaitObject(*(p)) - -WRes Thread_Wait_Close(CThread *p) -{ - WRes res = Thread_Wait(p); - WRes res2 = Thread_Close(p); - return (res != 0 ? res : res2); -} - -WRes Thread_Create(CThread *p, THREAD_FUNC_TYPE func, LPVOID param) -{ - /* Windows Me/98/95: threadId parameter may not be NULL in _beginthreadex/CreateThread functions */ - - #ifdef USE_THREADS_CreateThread - - DWORD threadId; - *p = CreateThread(NULL, 0, func, param, 0, &threadId); - - #else - - unsigned threadId; - *p = (HANDLE)(_beginthreadex(NULL, 0, func, param, 0, &threadId)); - - #endif - - /* maybe we must use errno here, but probably GetLastError() is also OK. */ - return HandleToWRes(*p); -} - - -WRes Thread_Create_With_Affinity(CThread *p, THREAD_FUNC_TYPE func, LPVOID param, CAffinityMask affinity) -{ - #ifdef USE_THREADS_CreateThread - - UNUSED_VAR(affinity) - return Thread_Create(p, func, param); - - #else - - /* Windows Me/98/95: threadId parameter may not be NULL in _beginthreadex/CreateThread functions */ - HANDLE h; - WRes wres; - unsigned threadId; - h = (HANDLE)(_beginthreadex(NULL, 0, func, param, CREATE_SUSPENDED, &threadId)); - *p = h; - wres = HandleToWRes(h); - if (h) - { - { - // DWORD_PTR prevMask = - SetThreadAffinityMask(h, (DWORD_PTR)affinity); - /* - if (prevMask == 0) - { - // affinity change is non-critical error, so we can ignore it - // wres = GetError(); - } - */ - } - { - DWORD prevSuspendCount = ResumeThread(h); - /* ResumeThread() returns: - 0 : was_not_suspended - 1 : was_resumed - -1 : error - */ - if (prevSuspendCount == (DWORD)-1) - wres = GetError(); - } - } - - /* maybe we must use errno here, but probably GetLastError() is also OK. */ - return wres; - - #endif -} - - -static WRes Event_Create(CEvent *p, BOOL manualReset, int signaled) -{ - *p = CreateEvent(NULL, manualReset, (signaled ? TRUE : FALSE), NULL); - return HandleToWRes(*p); -} - -WRes Event_Set(CEvent *p) { return BOOLToWRes(SetEvent(*p)); } -WRes Event_Reset(CEvent *p) { return BOOLToWRes(ResetEvent(*p)); } - -WRes ManualResetEvent_Create(CManualResetEvent *p, int signaled) { return Event_Create(p, TRUE, signaled); } -WRes AutoResetEvent_Create(CAutoResetEvent *p, int signaled) { return Event_Create(p, FALSE, signaled); } -WRes ManualResetEvent_CreateNotSignaled(CManualResetEvent *p) { return ManualResetEvent_Create(p, 0); } -WRes AutoResetEvent_CreateNotSignaled(CAutoResetEvent *p) { return AutoResetEvent_Create(p, 0); } - - -WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount) -{ - // negative ((LONG)maxCount) is not supported in WIN32::CreateSemaphore() - *p = CreateSemaphore(NULL, (LONG)initCount, (LONG)maxCount, NULL); - return HandleToWRes(*p); -} - -WRes Semaphore_OptCreateInit(CSemaphore *p, UInt32 initCount, UInt32 maxCount) -{ - // if (Semaphore_IsCreated(p)) - { - WRes wres = Semaphore_Close(p); - if (wres != 0) - return wres; - } - return Semaphore_Create(p, initCount, maxCount); -} - -static WRes Semaphore_Release(CSemaphore *p, LONG releaseCount, LONG *previousCount) - { return BOOLToWRes(ReleaseSemaphore(*p, releaseCount, previousCount)); } -WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 num) - { return Semaphore_Release(p, (LONG)num, NULL); } -WRes Semaphore_Release1(CSemaphore *p) { return Semaphore_ReleaseN(p, 1); } - -WRes CriticalSection_Init(CCriticalSection *p) -{ - /* InitializeCriticalSection() can raise exception: - Windows XP, 2003 : can raise a STATUS_NO_MEMORY exception - Windows Vista+ : no exceptions */ - #ifdef _MSC_VER - __try - #endif - { - InitializeCriticalSection(p); - /* InitializeCriticalSectionAndSpinCount(p, 0); */ - } - #ifdef _MSC_VER - __except (EXCEPTION_EXECUTE_HANDLER) { return ERROR_NOT_ENOUGH_MEMORY; } - #endif - return 0; -} - - - - -#else // _WIN32 - -// ---------- POSIX ---------- - -#ifndef __APPLE__ -#ifndef _7ZIP_AFFINITY_DISABLE -// _GNU_SOURCE can be required for pthread_setaffinity_np() / CPU_ZERO / CPU_SET -#define _GNU_SOURCE -#endif -#endif - -#include "Threads.h" - -#include -#include -#include -#ifdef _7ZIP_AFFINITY_SUPPORTED -// #include -#endif - - -// #include -// #define PRF(p) p -#define PRF(p) - -#define Print(s) PRF(printf("\n%s\n", s)) - -// #include - -WRes Thread_Create_With_CpuSet(CThread *p, THREAD_FUNC_TYPE func, LPVOID param, const CCpuSet *cpuSet) -{ - // new thread in Posix probably inherits affinity from parrent thread - Print("Thread_Create_With_CpuSet"); - - pthread_attr_t attr; - int ret; - // int ret2; - - p->_created = 0; - - RINOK(pthread_attr_init(&attr)); - - ret = pthread_attr_setdetachstate(&attr, PTHREAD_CREATE_JOINABLE); - - if (!ret) - { - if (cpuSet) - { - #ifdef _7ZIP_AFFINITY_SUPPORTED - - /* - printf("\n affinity :"); - unsigned i; - for (i = 0; i < sizeof(*cpuSet) && i < 8; i++) - { - Byte b = *((const Byte *)cpuSet + i); - char temp[32]; - #define GET_HEX_CHAR(t) ((char)(((t < 10) ? ('0' + t) : ('A' + (t - 10))))) - temp[0] = GET_HEX_CHAR((b & 0xF)); - temp[1] = GET_HEX_CHAR((b >> 4)); - // temp[0] = GET_HEX_CHAR((b >> 4)); // big-endian - // temp[1] = GET_HEX_CHAR((b & 0xF)); // big-endian - temp[2] = 0; - printf("%s", temp); - } - printf("\n"); - */ - - // ret2 = - pthread_attr_setaffinity_np(&attr, sizeof(*cpuSet), cpuSet); - // if (ret2) ret = ret2; - #endif - } - - ret = pthread_create(&p->_tid, &attr, func, param); - - if (!ret) - { - p->_created = 1; - /* - if (cpuSet) - { - // ret2 = - pthread_setaffinity_np(p->_tid, sizeof(*cpuSet), cpuSet); - // if (ret2) ret = ret2; - } - */ - } - } - // ret2 = - pthread_attr_destroy(&attr); - // if (ret2 != 0) ret = ret2; - return ret; -} - - -WRes Thread_Create(CThread *p, THREAD_FUNC_TYPE func, LPVOID param) -{ - return Thread_Create_With_CpuSet(p, func, param, NULL); -} - - -WRes Thread_Create_With_Affinity(CThread *p, THREAD_FUNC_TYPE func, LPVOID param, CAffinityMask affinity) -{ - Print("Thread_Create_WithAffinity"); - CCpuSet cs; - unsigned i; - CpuSet_Zero(&cs); - for (i = 0; i < sizeof(affinity) * 8; i++) - { - if (affinity == 0) - break; - if (affinity & 1) - { - CpuSet_Set(&cs, i); - } - affinity >>= 1; - } - return Thread_Create_With_CpuSet(p, func, param, &cs); -} - - -WRes Thread_Close(CThread *p) -{ - // Print("Thread_Close"); - int ret; - if (!p->_created) - return 0; - - ret = pthread_detach(p->_tid); - p->_tid = 0; - p->_created = 0; - return ret; -} - - -WRes Thread_Wait_Close(CThread *p) -{ - // Print("Thread_Wait_Close"); - void *thread_return; - int ret; - if (!p->_created) - return EINVAL; - - ret = pthread_join(p->_tid, &thread_return); - // probably we can't use that (_tid) after pthread_join(), so we close thread here - p->_created = 0; - p->_tid = 0; - return ret; -} - - - -static WRes Event_Create(CEvent *p, int manualReset, int signaled) -{ - RINOK(pthread_mutex_init(&p->_mutex, NULL)); - RINOK(pthread_cond_init(&p->_cond, NULL)); - p->_manual_reset = manualReset; - p->_state = (signaled ? True : False); - p->_created = 1; - return 0; -} - -WRes ManualResetEvent_Create(CManualResetEvent *p, int signaled) - { return Event_Create(p, True, signaled); } -WRes ManualResetEvent_CreateNotSignaled(CManualResetEvent *p) - { return ManualResetEvent_Create(p, 0); } -WRes AutoResetEvent_Create(CAutoResetEvent *p, int signaled) - { return Event_Create(p, False, signaled); } -WRes AutoResetEvent_CreateNotSignaled(CAutoResetEvent *p) - { return AutoResetEvent_Create(p, 0); } - - -WRes Event_Set(CEvent *p) -{ - RINOK(pthread_mutex_lock(&p->_mutex)); - p->_state = True; - int res1 = pthread_cond_broadcast(&p->_cond); - int res2 = pthread_mutex_unlock(&p->_mutex); - return (res2 ? res2 : res1); -} - -WRes Event_Reset(CEvent *p) -{ - RINOK(pthread_mutex_lock(&p->_mutex)); - p->_state = False; - return pthread_mutex_unlock(&p->_mutex); -} - -WRes Event_Wait(CEvent *p) -{ - RINOK(pthread_mutex_lock(&p->_mutex)); - while (p->_state == False) - { - // ETIMEDOUT - // ret = - pthread_cond_wait(&p->_cond, &p->_mutex); - // if (ret != 0) break; - } - if (p->_manual_reset == False) - { - p->_state = False; - } - return pthread_mutex_unlock(&p->_mutex); -} - -WRes Event_Close(CEvent *p) -{ - if (!p->_created) - return 0; - p->_created = 0; - { - int res1 = pthread_mutex_destroy(&p->_mutex); - int res2 = pthread_cond_destroy(&p->_cond); - return (res1 ? res1 : res2); - } -} - - -WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount) -{ - if (initCount > maxCount || maxCount < 1) - return EINVAL; - RINOK(pthread_mutex_init(&p->_mutex, NULL)); - RINOK(pthread_cond_init(&p->_cond, NULL)); - p->_count = initCount; - p->_maxCount = maxCount; - p->_created = 1; - return 0; -} - - -WRes Semaphore_OptCreateInit(CSemaphore *p, UInt32 initCount, UInt32 maxCount) -{ - if (Semaphore_IsCreated(p)) - { - /* - WRes wres = Semaphore_Close(p); - if (wres != 0) - return wres; - */ - if (initCount > maxCount || maxCount < 1) - return EINVAL; - // return EINVAL; // for debug - p->_count = initCount; - p->_maxCount = maxCount; - return 0; - } - return Semaphore_Create(p, initCount, maxCount); -} - - -WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 releaseCount) -{ - UInt32 newCount; - int ret; - - if (releaseCount < 1) - return EINVAL; - - RINOK(pthread_mutex_lock(&p->_mutex)); - - newCount = p->_count + releaseCount; - if (newCount > p->_maxCount) - ret = ERROR_TOO_MANY_POSTS; // EINVAL; - else - { - p->_count = newCount; - ret = pthread_cond_broadcast(&p->_cond); - } - RINOK(pthread_mutex_unlock(&p->_mutex)); - return ret; -} - -WRes Semaphore_Wait(CSemaphore *p) -{ - RINOK(pthread_mutex_lock(&p->_mutex)); - while (p->_count < 1) - { - pthread_cond_wait(&p->_cond, &p->_mutex); - } - p->_count--; - return pthread_mutex_unlock(&p->_mutex); -} - -WRes Semaphore_Close(CSemaphore *p) -{ - if (!p->_created) - return 0; - p->_created = 0; - { - int res1 = pthread_mutex_destroy(&p->_mutex); - int res2 = pthread_cond_destroy(&p->_cond); - return (res1 ? res1 : res2); - } -} - - - -WRes CriticalSection_Init(CCriticalSection *p) -{ - // Print("CriticalSection_Init"); - if (!p) - return EINTR; - return pthread_mutex_init(&p->_mutex, NULL); -} - -void CriticalSection_Enter(CCriticalSection *p) -{ - // Print("CriticalSection_Enter"); - if (p) - { - // int ret = - pthread_mutex_lock(&p->_mutex); - } -} - -void CriticalSection_Leave(CCriticalSection *p) -{ - // Print("CriticalSection_Leave"); - if (p) - { - // int ret = - pthread_mutex_unlock(&p->_mutex); - } -} - -void CriticalSection_Delete(CCriticalSection *p) -{ - // Print("CriticalSection_Delete"); - if (p) - { - // int ret = - pthread_mutex_destroy(&p->_mutex); - } -} - -LONG InterlockedIncrement(LONG volatile *addend) -{ - // Print("InterlockedIncrement"); - #ifdef USE_HACK_UNSAFE_ATOMIC - LONG val = *addend + 1; - *addend = val; - return val; - #else - return __sync_add_and_fetch(addend, 1); - #endif -} - -#endif // _WIN32 +/* Threads.c -- multithreading library +2021-12-21 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#ifdef _WIN32 + +#ifndef USE_THREADS_CreateThread +#include +#endif + +#include "Threads.h" + +static WRes GetError() +{ + DWORD res = GetLastError(); + return res ? (WRes)res : 1; +} + +static WRes HandleToWRes(HANDLE h) { return (h != NULL) ? 0 : GetError(); } +static WRes BOOLToWRes(BOOL v) { return v ? 0 : GetError(); } + +WRes HandlePtr_Close(HANDLE *p) +{ + if (*p != NULL) + { + if (!CloseHandle(*p)) + return GetError(); + *p = NULL; + } + return 0; +} + +WRes Handle_WaitObject(HANDLE h) +{ + DWORD dw = WaitForSingleObject(h, INFINITE); + /* + (dw) result: + WAIT_OBJECT_0 // 0 + WAIT_ABANDONED // 0x00000080 : is not compatible with Win32 Error space + WAIT_TIMEOUT // 0x00000102 : is compatible with Win32 Error space + WAIT_FAILED // 0xFFFFFFFF + */ + if (dw == WAIT_FAILED) + { + dw = GetLastError(); + if (dw == 0) + return WAIT_FAILED; + } + return (WRes)dw; +} + +#define Thread_Wait(p) Handle_WaitObject(*(p)) + +WRes Thread_Wait_Close(CThread *p) +{ + WRes res = Thread_Wait(p); + WRes res2 = Thread_Close(p); + return (res != 0 ? res : res2); +} + +WRes Thread_Create(CThread *p, THREAD_FUNC_TYPE func, LPVOID param) +{ + /* Windows Me/98/95: threadId parameter may not be NULL in _beginthreadex/CreateThread functions */ + + #ifdef USE_THREADS_CreateThread + + DWORD threadId; + *p = CreateThread(NULL, 0, func, param, 0, &threadId); + + #else + + unsigned threadId; + *p = (HANDLE)(_beginthreadex(NULL, 0, func, param, 0, &threadId)); + + #endif + + /* maybe we must use errno here, but probably GetLastError() is also OK. */ + return HandleToWRes(*p); +} + + +WRes Thread_Create_With_Affinity(CThread *p, THREAD_FUNC_TYPE func, LPVOID param, CAffinityMask affinity) +{ + #ifdef USE_THREADS_CreateThread + + UNUSED_VAR(affinity) + return Thread_Create(p, func, param); + + #else + + /* Windows Me/98/95: threadId parameter may not be NULL in _beginthreadex/CreateThread functions */ + HANDLE h; + WRes wres; + unsigned threadId; + h = (HANDLE)(_beginthreadex(NULL, 0, func, param, CREATE_SUSPENDED, &threadId)); + *p = h; + wres = HandleToWRes(h); + if (h) + { + { + // DWORD_PTR prevMask = + SetThreadAffinityMask(h, (DWORD_PTR)affinity); + /* + if (prevMask == 0) + { + // affinity change is non-critical error, so we can ignore it + // wres = GetError(); + } + */ + } + { + DWORD prevSuspendCount = ResumeThread(h); + /* ResumeThread() returns: + 0 : was_not_suspended + 1 : was_resumed + -1 : error + */ + if (prevSuspendCount == (DWORD)-1) + wres = GetError(); + } + } + + /* maybe we must use errno here, but probably GetLastError() is also OK. */ + return wres; + + #endif +} + + +static WRes Event_Create(CEvent *p, BOOL manualReset, int signaled) +{ + *p = CreateEvent(NULL, manualReset, (signaled ? TRUE : FALSE), NULL); + return HandleToWRes(*p); +} + +WRes Event_Set(CEvent *p) { return BOOLToWRes(SetEvent(*p)); } +WRes Event_Reset(CEvent *p) { return BOOLToWRes(ResetEvent(*p)); } + +WRes ManualResetEvent_Create(CManualResetEvent *p, int signaled) { return Event_Create(p, TRUE, signaled); } +WRes AutoResetEvent_Create(CAutoResetEvent *p, int signaled) { return Event_Create(p, FALSE, signaled); } +WRes ManualResetEvent_CreateNotSignaled(CManualResetEvent *p) { return ManualResetEvent_Create(p, 0); } +WRes AutoResetEvent_CreateNotSignaled(CAutoResetEvent *p) { return AutoResetEvent_Create(p, 0); } + + +WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount) +{ + // negative ((LONG)maxCount) is not supported in WIN32::CreateSemaphore() + *p = CreateSemaphore(NULL, (LONG)initCount, (LONG)maxCount, NULL); + return HandleToWRes(*p); +} + +WRes Semaphore_OptCreateInit(CSemaphore *p, UInt32 initCount, UInt32 maxCount) +{ + // if (Semaphore_IsCreated(p)) + { + WRes wres = Semaphore_Close(p); + if (wres != 0) + return wres; + } + return Semaphore_Create(p, initCount, maxCount); +} + +static WRes Semaphore_Release(CSemaphore *p, LONG releaseCount, LONG *previousCount) + { return BOOLToWRes(ReleaseSemaphore(*p, releaseCount, previousCount)); } +WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 num) + { return Semaphore_Release(p, (LONG)num, NULL); } +WRes Semaphore_Release1(CSemaphore *p) { return Semaphore_ReleaseN(p, 1); } + +WRes CriticalSection_Init(CCriticalSection *p) +{ + /* InitializeCriticalSection() can raise exception: + Windows XP, 2003 : can raise a STATUS_NO_MEMORY exception + Windows Vista+ : no exceptions */ + #ifdef _MSC_VER + __try + #endif + { + InitializeCriticalSection(p); + /* InitializeCriticalSectionAndSpinCount(p, 0); */ + } + #ifdef _MSC_VER + __except (EXCEPTION_EXECUTE_HANDLER) { return ERROR_NOT_ENOUGH_MEMORY; } + #endif + return 0; +} + + + + +#else // _WIN32 + +// ---------- POSIX ---------- + +#ifndef __APPLE__ +#ifndef _7ZIP_AFFINITY_DISABLE +// _GNU_SOURCE can be required for pthread_setaffinity_np() / CPU_ZERO / CPU_SET +#define _GNU_SOURCE +#endif +#endif + +#include "Threads.h" + +#include +#include +#include +#ifdef _7ZIP_AFFINITY_SUPPORTED +// #include +#endif + + +// #include +// #define PRF(p) p +#define PRF(p) + +#define Print(s) PRF(printf("\n%s\n", s)) + +// #include + +WRes Thread_Create_With_CpuSet(CThread *p, THREAD_FUNC_TYPE func, LPVOID param, const CCpuSet *cpuSet) +{ + // new thread in Posix probably inherits affinity from parrent thread + Print("Thread_Create_With_CpuSet"); + + pthread_attr_t attr; + int ret; + // int ret2; + + p->_created = 0; + + RINOK(pthread_attr_init(&attr)); + + ret = pthread_attr_setdetachstate(&attr, PTHREAD_CREATE_JOINABLE); + + if (!ret) + { + if (cpuSet) + { + #ifdef _7ZIP_AFFINITY_SUPPORTED + + /* + printf("\n affinity :"); + unsigned i; + for (i = 0; i < sizeof(*cpuSet) && i < 8; i++) + { + Byte b = *((const Byte *)cpuSet + i); + char temp[32]; + #define GET_HEX_CHAR(t) ((char)(((t < 10) ? ('0' + t) : ('A' + (t - 10))))) + temp[0] = GET_HEX_CHAR((b & 0xF)); + temp[1] = GET_HEX_CHAR((b >> 4)); + // temp[0] = GET_HEX_CHAR((b >> 4)); // big-endian + // temp[1] = GET_HEX_CHAR((b & 0xF)); // big-endian + temp[2] = 0; + printf("%s", temp); + } + printf("\n"); + */ + + // ret2 = + pthread_attr_setaffinity_np(&attr, sizeof(*cpuSet), cpuSet); + // if (ret2) ret = ret2; + #endif + } + + ret = pthread_create(&p->_tid, &attr, func, param); + + if (!ret) + { + p->_created = 1; + /* + if (cpuSet) + { + // ret2 = + pthread_setaffinity_np(p->_tid, sizeof(*cpuSet), cpuSet); + // if (ret2) ret = ret2; + } + */ + } + } + // ret2 = + pthread_attr_destroy(&attr); + // if (ret2 != 0) ret = ret2; + return ret; +} + + +WRes Thread_Create(CThread *p, THREAD_FUNC_TYPE func, LPVOID param) +{ + return Thread_Create_With_CpuSet(p, func, param, NULL); +} + + +WRes Thread_Create_With_Affinity(CThread *p, THREAD_FUNC_TYPE func, LPVOID param, CAffinityMask affinity) +{ + Print("Thread_Create_WithAffinity"); + CCpuSet cs; + unsigned i; + CpuSet_Zero(&cs); + for (i = 0; i < sizeof(affinity) * 8; i++) + { + if (affinity == 0) + break; + if (affinity & 1) + { + CpuSet_Set(&cs, i); + } + affinity >>= 1; + } + return Thread_Create_With_CpuSet(p, func, param, &cs); +} + + +WRes Thread_Close(CThread *p) +{ + // Print("Thread_Close"); + int ret; + if (!p->_created) + return 0; + + ret = pthread_detach(p->_tid); + p->_tid = 0; + p->_created = 0; + return ret; +} + + +WRes Thread_Wait_Close(CThread *p) +{ + // Print("Thread_Wait_Close"); + void *thread_return; + int ret; + if (!p->_created) + return EINVAL; + + ret = pthread_join(p->_tid, &thread_return); + // probably we can't use that (_tid) after pthread_join(), so we close thread here + p->_created = 0; + p->_tid = 0; + return ret; +} + + + +static WRes Event_Create(CEvent *p, int manualReset, int signaled) +{ + RINOK(pthread_mutex_init(&p->_mutex, NULL)); + RINOK(pthread_cond_init(&p->_cond, NULL)); + p->_manual_reset = manualReset; + p->_state = (signaled ? True : False); + p->_created = 1; + return 0; +} + +WRes ManualResetEvent_Create(CManualResetEvent *p, int signaled) + { return Event_Create(p, True, signaled); } +WRes ManualResetEvent_CreateNotSignaled(CManualResetEvent *p) + { return ManualResetEvent_Create(p, 0); } +WRes AutoResetEvent_Create(CAutoResetEvent *p, int signaled) + { return Event_Create(p, False, signaled); } +WRes AutoResetEvent_CreateNotSignaled(CAutoResetEvent *p) + { return AutoResetEvent_Create(p, 0); } + + +WRes Event_Set(CEvent *p) +{ + RINOK(pthread_mutex_lock(&p->_mutex)); + p->_state = True; + int res1 = pthread_cond_broadcast(&p->_cond); + int res2 = pthread_mutex_unlock(&p->_mutex); + return (res2 ? res2 : res1); +} + +WRes Event_Reset(CEvent *p) +{ + RINOK(pthread_mutex_lock(&p->_mutex)); + p->_state = False; + return pthread_mutex_unlock(&p->_mutex); +} + +WRes Event_Wait(CEvent *p) +{ + RINOK(pthread_mutex_lock(&p->_mutex)); + while (p->_state == False) + { + // ETIMEDOUT + // ret = + pthread_cond_wait(&p->_cond, &p->_mutex); + // if (ret != 0) break; + } + if (p->_manual_reset == False) + { + p->_state = False; + } + return pthread_mutex_unlock(&p->_mutex); +} + +WRes Event_Close(CEvent *p) +{ + if (!p->_created) + return 0; + p->_created = 0; + { + int res1 = pthread_mutex_destroy(&p->_mutex); + int res2 = pthread_cond_destroy(&p->_cond); + return (res1 ? res1 : res2); + } +} + + +WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount) +{ + if (initCount > maxCount || maxCount < 1) + return EINVAL; + RINOK(pthread_mutex_init(&p->_mutex, NULL)); + RINOK(pthread_cond_init(&p->_cond, NULL)); + p->_count = initCount; + p->_maxCount = maxCount; + p->_created = 1; + return 0; +} + + +WRes Semaphore_OptCreateInit(CSemaphore *p, UInt32 initCount, UInt32 maxCount) +{ + if (Semaphore_IsCreated(p)) + { + /* + WRes wres = Semaphore_Close(p); + if (wres != 0) + return wres; + */ + if (initCount > maxCount || maxCount < 1) + return EINVAL; + // return EINVAL; // for debug + p->_count = initCount; + p->_maxCount = maxCount; + return 0; + } + return Semaphore_Create(p, initCount, maxCount); +} + + +WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 releaseCount) +{ + UInt32 newCount; + int ret; + + if (releaseCount < 1) + return EINVAL; + + RINOK(pthread_mutex_lock(&p->_mutex)); + + newCount = p->_count + releaseCount; + if (newCount > p->_maxCount) + ret = ERROR_TOO_MANY_POSTS; // EINVAL; + else + { + p->_count = newCount; + ret = pthread_cond_broadcast(&p->_cond); + } + RINOK(pthread_mutex_unlock(&p->_mutex)); + return ret; +} + +WRes Semaphore_Wait(CSemaphore *p) +{ + RINOK(pthread_mutex_lock(&p->_mutex)); + while (p->_count < 1) + { + pthread_cond_wait(&p->_cond, &p->_mutex); + } + p->_count--; + return pthread_mutex_unlock(&p->_mutex); +} + +WRes Semaphore_Close(CSemaphore *p) +{ + if (!p->_created) + return 0; + p->_created = 0; + { + int res1 = pthread_mutex_destroy(&p->_mutex); + int res2 = pthread_cond_destroy(&p->_cond); + return (res1 ? res1 : res2); + } +} + + + +WRes CriticalSection_Init(CCriticalSection *p) +{ + // Print("CriticalSection_Init"); + if (!p) + return EINTR; + return pthread_mutex_init(&p->_mutex, NULL); +} + +void CriticalSection_Enter(CCriticalSection *p) +{ + // Print("CriticalSection_Enter"); + if (p) + { + // int ret = + pthread_mutex_lock(&p->_mutex); + } +} + +void CriticalSection_Leave(CCriticalSection *p) +{ + // Print("CriticalSection_Leave"); + if (p) + { + // int ret = + pthread_mutex_unlock(&p->_mutex); + } +} + +void CriticalSection_Delete(CCriticalSection *p) +{ + // Print("CriticalSection_Delete"); + if (p) + { + // int ret = + pthread_mutex_destroy(&p->_mutex); + } +} + +LONG InterlockedIncrement(LONG volatile *addend) +{ + // Print("InterlockedIncrement"); + #ifdef USE_HACK_UNSAFE_ATOMIC + LONG val = *addend + 1; + *addend = val; + return val; + #else + return __sync_add_and_fetch(addend, 1); + #endif +} + +#endif // _WIN32 diff --git a/deps/LZMA-SDK/C/Threads.h b/deps/LZMA-SDK/C/Threads.h index 71972558d..ceb844a23 100644 --- a/deps/LZMA-SDK/C/Threads.h +++ b/deps/LZMA-SDK/C/Threads.h @@ -1,232 +1,232 @@ -/* Threads.h -- multithreading library -2021-12-21 : Igor Pavlov : Public domain */ - -#ifndef __7Z_THREADS_H -#define __7Z_THREADS_H - -#ifdef _WIN32 -#include -#else - -#if defined(__linux__) -#if !defined(__APPLE__) && !defined(_AIX) && !defined(__ANDROID__) -#ifndef _7ZIP_AFFINITY_DISABLE -#define _7ZIP_AFFINITY_SUPPORTED -// #pragma message(" ==== _7ZIP_AFFINITY_SUPPORTED") -// #define _GNU_SOURCE -#endif -#endif -#endif - -#include - -#endif - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -#ifdef _WIN32 - -WRes HandlePtr_Close(HANDLE *h); -WRes Handle_WaitObject(HANDLE h); - -typedef HANDLE CThread; - -#define Thread_Construct(p) { *(p) = NULL; } -#define Thread_WasCreated(p) (*(p) != NULL) -#define Thread_Close(p) HandlePtr_Close(p) -// #define Thread_Wait(p) Handle_WaitObject(*(p)) - -#ifdef UNDER_CE - // if (USE_THREADS_CreateThread is defined), we use _beginthreadex() - // if (USE_THREADS_CreateThread is not definned), we use CreateThread() - #define USE_THREADS_CreateThread -#endif - -typedef - #ifdef USE_THREADS_CreateThread - DWORD - #else - unsigned - #endif - THREAD_FUNC_RET_TYPE; - -typedef DWORD_PTR CAffinityMask; -typedef DWORD_PTR CCpuSet; - -#define CpuSet_Zero(p) { *(p) = 0; } -#define CpuSet_Set(p, cpu) { *(p) |= ((DWORD_PTR)1 << (cpu)); } - -#else // _WIN32 - -typedef struct _CThread -{ - pthread_t _tid; - int _created; -} CThread; - -#define Thread_Construct(p) { (p)->_tid = 0; (p)->_created = 0; } -#define Thread_WasCreated(p) ((p)->_created != 0) -WRes Thread_Close(CThread *p); -// #define Thread_Wait Thread_Wait_Close - -typedef void * THREAD_FUNC_RET_TYPE; - -typedef UInt64 CAffinityMask; - -#ifdef _7ZIP_AFFINITY_SUPPORTED - -typedef cpu_set_t CCpuSet; -#define CpuSet_Zero(p) CPU_ZERO(p) -#define CpuSet_Set(p, cpu) CPU_SET(cpu, p) -#define CpuSet_IsSet(p, cpu) CPU_ISSET(cpu, p) - -#else - -typedef UInt64 CCpuSet; -#define CpuSet_Zero(p) { *(p) = 0; } -#define CpuSet_Set(p, cpu) { *(p) |= ((UInt64)1 << (cpu)); } -#define CpuSet_IsSet(p, cpu) ((*(p) & ((UInt64)1 << (cpu))) != 0) - -#endif - - -#endif // _WIN32 - - -#define THREAD_FUNC_CALL_TYPE MY_STD_CALL - -#if defined(_WIN32) && defined(__GNUC__) -/* GCC compiler for x86 32-bit uses the rule: - the stack is 16-byte aligned before CALL instruction for function calling. - But only root function main() contains instructions that - set 16-byte alignment for stack pointer. And another functions - just keep alignment, if it was set in some parent function. - - The problem: - if we create new thread in MinGW (GCC) 32-bit x86 via _beginthreadex() or CreateThread(), - the root function of thread doesn't set 16-byte alignment. - And stack frames in all child functions also will be unaligned in that case. - - Here we set (force_align_arg_pointer) attribute for root function of new thread. - Do we need (force_align_arg_pointer) also for another systems? */ - - #define THREAD_FUNC_ATTRIB_ALIGN_ARG __attribute__((force_align_arg_pointer)) - // #define THREAD_FUNC_ATTRIB_ALIGN_ARG // for debug : bad alignment in SSE functions -#else - #define THREAD_FUNC_ATTRIB_ALIGN_ARG -#endif - -#define THREAD_FUNC_DECL THREAD_FUNC_ATTRIB_ALIGN_ARG THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE - -typedef THREAD_FUNC_RET_TYPE (THREAD_FUNC_CALL_TYPE * THREAD_FUNC_TYPE)(void *); -WRes Thread_Create(CThread *p, THREAD_FUNC_TYPE func, LPVOID param); -WRes Thread_Create_With_Affinity(CThread *p, THREAD_FUNC_TYPE func, LPVOID param, CAffinityMask affinity); -WRes Thread_Wait_Close(CThread *p); - -#ifdef _WIN32 -#define Thread_Create_With_CpuSet(p, func, param, cs) \ - Thread_Create_With_Affinity(p, func, param, *cs) -#else -WRes Thread_Create_With_CpuSet(CThread *p, THREAD_FUNC_TYPE func, LPVOID param, const CCpuSet *cpuSet); -#endif - - -#ifdef _WIN32 - -typedef HANDLE CEvent; -typedef CEvent CAutoResetEvent; -typedef CEvent CManualResetEvent; -#define Event_Construct(p) *(p) = NULL -#define Event_IsCreated(p) (*(p) != NULL) -#define Event_Close(p) HandlePtr_Close(p) -#define Event_Wait(p) Handle_WaitObject(*(p)) -WRes Event_Set(CEvent *p); -WRes Event_Reset(CEvent *p); -WRes ManualResetEvent_Create(CManualResetEvent *p, int signaled); -WRes ManualResetEvent_CreateNotSignaled(CManualResetEvent *p); -WRes AutoResetEvent_Create(CAutoResetEvent *p, int signaled); -WRes AutoResetEvent_CreateNotSignaled(CAutoResetEvent *p); - -typedef HANDLE CSemaphore; -#define Semaphore_Construct(p) *(p) = NULL -#define Semaphore_IsCreated(p) (*(p) != NULL) -#define Semaphore_Close(p) HandlePtr_Close(p) -#define Semaphore_Wait(p) Handle_WaitObject(*(p)) -WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount); -WRes Semaphore_OptCreateInit(CSemaphore *p, UInt32 initCount, UInt32 maxCount); -WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 num); -WRes Semaphore_Release1(CSemaphore *p); - -typedef CRITICAL_SECTION CCriticalSection; -WRes CriticalSection_Init(CCriticalSection *p); -#define CriticalSection_Delete(p) DeleteCriticalSection(p) -#define CriticalSection_Enter(p) EnterCriticalSection(p) -#define CriticalSection_Leave(p) LeaveCriticalSection(p) - - -#else // _WIN32 - -typedef struct _CEvent -{ - int _created; - int _manual_reset; - int _state; - pthread_mutex_t _mutex; - pthread_cond_t _cond; -} CEvent; - -typedef CEvent CAutoResetEvent; -typedef CEvent CManualResetEvent; - -#define Event_Construct(p) (p)->_created = 0 -#define Event_IsCreated(p) ((p)->_created) - -WRes ManualResetEvent_Create(CManualResetEvent *p, int signaled); -WRes ManualResetEvent_CreateNotSignaled(CManualResetEvent *p); -WRes AutoResetEvent_Create(CAutoResetEvent *p, int signaled); -WRes AutoResetEvent_CreateNotSignaled(CAutoResetEvent *p); -WRes Event_Set(CEvent *p); -WRes Event_Reset(CEvent *p); -WRes Event_Wait(CEvent *p); -WRes Event_Close(CEvent *p); - - -typedef struct _CSemaphore -{ - int _created; - UInt32 _count; - UInt32 _maxCount; - pthread_mutex_t _mutex; - pthread_cond_t _cond; -} CSemaphore; - -#define Semaphore_Construct(p) (p)->_created = 0 -#define Semaphore_IsCreated(p) ((p)->_created) - -WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount); -WRes Semaphore_OptCreateInit(CSemaphore *p, UInt32 initCount, UInt32 maxCount); -WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 num); -#define Semaphore_Release1(p) Semaphore_ReleaseN(p, 1) -WRes Semaphore_Wait(CSemaphore *p); -WRes Semaphore_Close(CSemaphore *p); - - -typedef struct _CCriticalSection -{ - pthread_mutex_t _mutex; -} CCriticalSection; - -WRes CriticalSection_Init(CCriticalSection *p); -void CriticalSection_Delete(CCriticalSection *cs); -void CriticalSection_Enter(CCriticalSection *cs); -void CriticalSection_Leave(CCriticalSection *cs); - -LONG InterlockedIncrement(LONG volatile *addend); - -#endif // _WIN32 - -EXTERN_C_END - -#endif +/* Threads.h -- multithreading library +2021-12-21 : Igor Pavlov : Public domain */ + +#ifndef __7Z_THREADS_H +#define __7Z_THREADS_H + +#ifdef _WIN32 +#include +#else + +#if defined(__linux__) +#if !defined(__APPLE__) && !defined(_AIX) && !defined(__ANDROID__) +#ifndef _7ZIP_AFFINITY_DISABLE +#define _7ZIP_AFFINITY_SUPPORTED +// #pragma message(" ==== _7ZIP_AFFINITY_SUPPORTED") +// #define _GNU_SOURCE +#endif +#endif +#endif + +#include + +#endif + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#ifdef _WIN32 + +WRes HandlePtr_Close(HANDLE *h); +WRes Handle_WaitObject(HANDLE h); + +typedef HANDLE CThread; + +#define Thread_Construct(p) { *(p) = NULL; } +#define Thread_WasCreated(p) (*(p) != NULL) +#define Thread_Close(p) HandlePtr_Close(p) +// #define Thread_Wait(p) Handle_WaitObject(*(p)) + +#ifdef UNDER_CE + // if (USE_THREADS_CreateThread is defined), we use _beginthreadex() + // if (USE_THREADS_CreateThread is not definned), we use CreateThread() + #define USE_THREADS_CreateThread +#endif + +typedef + #ifdef USE_THREADS_CreateThread + DWORD + #else + unsigned + #endif + THREAD_FUNC_RET_TYPE; + +typedef DWORD_PTR CAffinityMask; +typedef DWORD_PTR CCpuSet; + +#define CpuSet_Zero(p) { *(p) = 0; } +#define CpuSet_Set(p, cpu) { *(p) |= ((DWORD_PTR)1 << (cpu)); } + +#else // _WIN32 + +typedef struct _CThread +{ + pthread_t _tid; + int _created; +} CThread; + +#define Thread_Construct(p) { (p)->_tid = 0; (p)->_created = 0; } +#define Thread_WasCreated(p) ((p)->_created != 0) +WRes Thread_Close(CThread *p); +// #define Thread_Wait Thread_Wait_Close + +typedef void * THREAD_FUNC_RET_TYPE; + +typedef UInt64 CAffinityMask; + +#ifdef _7ZIP_AFFINITY_SUPPORTED + +typedef cpu_set_t CCpuSet; +#define CpuSet_Zero(p) CPU_ZERO(p) +#define CpuSet_Set(p, cpu) CPU_SET(cpu, p) +#define CpuSet_IsSet(p, cpu) CPU_ISSET(cpu, p) + +#else + +typedef UInt64 CCpuSet; +#define CpuSet_Zero(p) { *(p) = 0; } +#define CpuSet_Set(p, cpu) { *(p) |= ((UInt64)1 << (cpu)); } +#define CpuSet_IsSet(p, cpu) ((*(p) & ((UInt64)1 << (cpu))) != 0) + +#endif + + +#endif // _WIN32 + + +#define THREAD_FUNC_CALL_TYPE MY_STD_CALL + +#if defined(_WIN32) && defined(__GNUC__) +/* GCC compiler for x86 32-bit uses the rule: + the stack is 16-byte aligned before CALL instruction for function calling. + But only root function main() contains instructions that + set 16-byte alignment for stack pointer. And another functions + just keep alignment, if it was set in some parent function. + + The problem: + if we create new thread in MinGW (GCC) 32-bit x86 via _beginthreadex() or CreateThread(), + the root function of thread doesn't set 16-byte alignment. + And stack frames in all child functions also will be unaligned in that case. + + Here we set (force_align_arg_pointer) attribute for root function of new thread. + Do we need (force_align_arg_pointer) also for another systems? */ + + #define THREAD_FUNC_ATTRIB_ALIGN_ARG __attribute__((force_align_arg_pointer)) + // #define THREAD_FUNC_ATTRIB_ALIGN_ARG // for debug : bad alignment in SSE functions +#else + #define THREAD_FUNC_ATTRIB_ALIGN_ARG +#endif + +#define THREAD_FUNC_DECL THREAD_FUNC_ATTRIB_ALIGN_ARG THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE + +typedef THREAD_FUNC_RET_TYPE (THREAD_FUNC_CALL_TYPE * THREAD_FUNC_TYPE)(void *); +WRes Thread_Create(CThread *p, THREAD_FUNC_TYPE func, LPVOID param); +WRes Thread_Create_With_Affinity(CThread *p, THREAD_FUNC_TYPE func, LPVOID param, CAffinityMask affinity); +WRes Thread_Wait_Close(CThread *p); + +#ifdef _WIN32 +#define Thread_Create_With_CpuSet(p, func, param, cs) \ + Thread_Create_With_Affinity(p, func, param, *cs) +#else +WRes Thread_Create_With_CpuSet(CThread *p, THREAD_FUNC_TYPE func, LPVOID param, const CCpuSet *cpuSet); +#endif + + +#ifdef _WIN32 + +typedef HANDLE CEvent; +typedef CEvent CAutoResetEvent; +typedef CEvent CManualResetEvent; +#define Event_Construct(p) *(p) = NULL +#define Event_IsCreated(p) (*(p) != NULL) +#define Event_Close(p) HandlePtr_Close(p) +#define Event_Wait(p) Handle_WaitObject(*(p)) +WRes Event_Set(CEvent *p); +WRes Event_Reset(CEvent *p); +WRes ManualResetEvent_Create(CManualResetEvent *p, int signaled); +WRes ManualResetEvent_CreateNotSignaled(CManualResetEvent *p); +WRes AutoResetEvent_Create(CAutoResetEvent *p, int signaled); +WRes AutoResetEvent_CreateNotSignaled(CAutoResetEvent *p); + +typedef HANDLE CSemaphore; +#define Semaphore_Construct(p) *(p) = NULL +#define Semaphore_IsCreated(p) (*(p) != NULL) +#define Semaphore_Close(p) HandlePtr_Close(p) +#define Semaphore_Wait(p) Handle_WaitObject(*(p)) +WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount); +WRes Semaphore_OptCreateInit(CSemaphore *p, UInt32 initCount, UInt32 maxCount); +WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 num); +WRes Semaphore_Release1(CSemaphore *p); + +typedef CRITICAL_SECTION CCriticalSection; +WRes CriticalSection_Init(CCriticalSection *p); +#define CriticalSection_Delete(p) DeleteCriticalSection(p) +#define CriticalSection_Enter(p) EnterCriticalSection(p) +#define CriticalSection_Leave(p) LeaveCriticalSection(p) + + +#else // _WIN32 + +typedef struct _CEvent +{ + int _created; + int _manual_reset; + int _state; + pthread_mutex_t _mutex; + pthread_cond_t _cond; +} CEvent; + +typedef CEvent CAutoResetEvent; +typedef CEvent CManualResetEvent; + +#define Event_Construct(p) (p)->_created = 0 +#define Event_IsCreated(p) ((p)->_created) + +WRes ManualResetEvent_Create(CManualResetEvent *p, int signaled); +WRes ManualResetEvent_CreateNotSignaled(CManualResetEvent *p); +WRes AutoResetEvent_Create(CAutoResetEvent *p, int signaled); +WRes AutoResetEvent_CreateNotSignaled(CAutoResetEvent *p); +WRes Event_Set(CEvent *p); +WRes Event_Reset(CEvent *p); +WRes Event_Wait(CEvent *p); +WRes Event_Close(CEvent *p); + + +typedef struct _CSemaphore +{ + int _created; + UInt32 _count; + UInt32 _maxCount; + pthread_mutex_t _mutex; + pthread_cond_t _cond; +} CSemaphore; + +#define Semaphore_Construct(p) (p)->_created = 0 +#define Semaphore_IsCreated(p) ((p)->_created) + +WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount); +WRes Semaphore_OptCreateInit(CSemaphore *p, UInt32 initCount, UInt32 maxCount); +WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 num); +#define Semaphore_Release1(p) Semaphore_ReleaseN(p, 1) +WRes Semaphore_Wait(CSemaphore *p); +WRes Semaphore_Close(CSemaphore *p); + + +typedef struct _CCriticalSection +{ + pthread_mutex_t _mutex; +} CCriticalSection; + +WRes CriticalSection_Init(CCriticalSection *p); +void CriticalSection_Delete(CCriticalSection *cs); +void CriticalSection_Enter(CCriticalSection *cs); +void CriticalSection_Leave(CCriticalSection *cs); + +LONG InterlockedIncrement(LONG volatile *addend); + +#endif // _WIN32 + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/Util/7z/7z.dsp b/deps/LZMA-SDK/C/Util/7z/7z.dsp index d3bf0fe00..be0f0a743 100644 --- a/deps/LZMA-SDK/C/Util/7z/7z.dsp +++ b/deps/LZMA-SDK/C/Util/7z/7z.dsp @@ -1,241 +1,241 @@ -# Microsoft Developer Studio Project File - Name="7z" - Package Owner=<4> -# Microsoft Developer Studio Generated Build File, Format Version 6.00 -# ** DO NOT EDIT ** - -# TARGTYPE "Win32 (x86) Console Application" 0x0103 - -CFG=7z - Win32 Debug -!MESSAGE This is not a valid makefile. To build this project using NMAKE, -!MESSAGE use the Export Makefile command and run -!MESSAGE -!MESSAGE NMAKE /f "7z.mak". -!MESSAGE -!MESSAGE You can specify a configuration when running NMAKE -!MESSAGE by defining the macro CFG on the command line. For example: -!MESSAGE -!MESSAGE NMAKE /f "7z.mak" CFG="7z - Win32 Debug" -!MESSAGE -!MESSAGE Possible choices for configuration are: -!MESSAGE -!MESSAGE "7z - Win32 Release" (based on "Win32 (x86) Console Application") -!MESSAGE "7z - Win32 Debug" (based on "Win32 (x86) Console Application") -!MESSAGE - -# Begin Project -# PROP AllowPerConfigDependencies 0 -# PROP Scc_ProjName "" -# PROP Scc_LocalPath "" -CPP=cl.exe -RSC=rc.exe - -!IF "$(CFG)" == "7z - Win32 Release" - -# PROP BASE Use_MFC 0 -# PROP BASE Use_Debug_Libraries 0 -# PROP BASE Output_Dir "Release" -# PROP BASE Intermediate_Dir "Release" -# PROP BASE Target_Dir "" -# PROP Use_MFC 0 -# PROP Use_Debug_Libraries 0 -# PROP Output_Dir "Release" -# PROP Intermediate_Dir "Release" -# PROP Ignore_Export_Lib 0 -# PROP Target_Dir "" -# ADD BASE CPP /nologo /W3 /GX /O2 /D "WIN32" /D "NDEBUG" /D "_CONSOLE" /D "_MBCS" /YX /FD /c -# ADD CPP /nologo /MD /W4 /WX /GX /O2 /D "NDEBUG" /D "WIN32" /D "_CONSOLE" /D "_UNICODE" /D "UNICODE" /FAcs /Yu"Precomp.h" /FD /c -# ADD BASE RSC /l 0x419 /d "NDEBUG" -# ADD RSC /l 0x419 /d "NDEBUG" -BSC32=bscmake.exe -# ADD BASE BSC32 /nologo -# ADD BSC32 /nologo -LINK32=link.exe -# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /machine:I386 -# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /machine:I386 /out:"c:\util\7zDec.exe" /opt:NOWIN98 -# SUBTRACT LINK32 /pdb:none - -!ELSEIF "$(CFG)" == "7z - Win32 Debug" - -# PROP BASE Use_MFC 0 -# PROP BASE Use_Debug_Libraries 1 -# PROP BASE Output_Dir "Debug" -# PROP BASE Intermediate_Dir "Debug" -# PROP BASE Target_Dir "" -# PROP Use_MFC 0 -# PROP Use_Debug_Libraries 1 -# PROP Output_Dir "Debug" -# PROP Intermediate_Dir "Debug" -# PROP Ignore_Export_Lib 0 -# PROP Target_Dir "" -# ADD BASE CPP /nologo /W3 /Gm /GX /ZI /Od /D "WIN32" /D "_DEBUG" /D "_CONSOLE" /D "_MBCS" /YX /FD /GZ /c -# ADD CPP /nologo /W4 /WX /Gm /GX /ZI /Od /D "_DEBUG" /D "_SZ_ALLOC_DEBUG2" /D "_SZ_NO_INT_64_A" /D "WIN32" /D "_CONSOLE" /D "_UNICODE" /D "UNICODE" /Yu"Precomp.h" /FD /GZ /c -# ADD BASE RSC /l 0x419 /d "_DEBUG" -# ADD RSC /l 0x419 /d "_DEBUG" -BSC32=bscmake.exe -# ADD BASE BSC32 /nologo -# ADD BSC32 /nologo -LINK32=link.exe -# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /debug /machine:I386 /pdbtype:sept -# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /debug /machine:I386 /out:"c:\util\7zDec.exe" /pdbtype:sept - -!ENDIF - -# Begin Target - -# Name "7z - Win32 Release" -# Name "7z - Win32 Debug" -# Begin Group "Common" - -# PROP Default_Filter "" -# Begin Source File - -SOURCE=..\..\7z.h -# End Source File -# Begin Source File - -SOURCE=..\..\7zAlloc.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zAlloc.h -# End Source File -# Begin Source File - -SOURCE=..\..\7zArcIn.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zBuf.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zBuf.h -# End Source File -# Begin Source File - -SOURCE=..\..\7zCrc.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zCrc.h -# End Source File -# Begin Source File - -SOURCE=..\..\7zCrcOpt.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zDec.c -# ADD CPP /D "_7ZIP_PPMD_SUPPPORT" -# End Source File -# Begin Source File - -SOURCE=..\..\7zFile.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zFile.h -# End Source File -# Begin Source File - -SOURCE=..\..\7zStream.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zTypes.h -# End Source File -# Begin Source File - -SOURCE=..\..\Bcj2.c -# End Source File -# Begin Source File - -SOURCE=..\..\Bcj2.h -# End Source File -# Begin Source File - -SOURCE=..\..\Bra.c -# End Source File -# Begin Source File - -SOURCE=..\..\Bra.h -# End Source File -# Begin Source File - -SOURCE=..\..\Bra86.c -# End Source File -# Begin Source File - -SOURCE=..\..\BraIA64.c -# End Source File -# Begin Source File - -SOURCE=..\..\CpuArch.c -# End Source File -# Begin Source File - -SOURCE=..\..\CpuArch.h -# End Source File -# Begin Source File - -SOURCE=..\..\Delta.c -# End Source File -# Begin Source File - -SOURCE=..\..\Delta.h -# End Source File -# Begin Source File - -SOURCE=..\..\Lzma2Dec.c -# End Source File -# Begin Source File - -SOURCE=..\..\Lzma2Dec.h -# End Source File -# Begin Source File - -SOURCE=..\..\LzmaDec.c -# End Source File -# Begin Source File - -SOURCE=..\..\LzmaDec.h -# End Source File -# Begin Source File - -SOURCE=..\..\Ppmd.h -# End Source File -# Begin Source File - -SOURCE=..\..\Ppmd7.c -# End Source File -# Begin Source File - -SOURCE=..\..\Ppmd7.h -# End Source File -# Begin Source File - -SOURCE=..\..\Ppmd7Dec.c -# End Source File -# End Group -# Begin Group "Spec" - -# PROP Default_Filter "" -# Begin Source File - -SOURCE=..\..\Compiler.h -# End Source File -# Begin Source File - -SOURCE=.\Precomp.c -# ADD CPP /Yc"Precomp.h" -# End Source File -# Begin Source File - -SOURCE=.\Precomp.h -# End Source File -# End Group -# Begin Source File - -SOURCE=.\7zMain.c -# End Source File -# End Target -# End Project +# Microsoft Developer Studio Project File - Name="7z" - Package Owner=<4> +# Microsoft Developer Studio Generated Build File, Format Version 6.00 +# ** DO NOT EDIT ** + +# TARGTYPE "Win32 (x86) Console Application" 0x0103 + +CFG=7z - Win32 Debug +!MESSAGE This is not a valid makefile. To build this project using NMAKE, +!MESSAGE use the Export Makefile command and run +!MESSAGE +!MESSAGE NMAKE /f "7z.mak". +!MESSAGE +!MESSAGE You can specify a configuration when running NMAKE +!MESSAGE by defining the macro CFG on the command line. For example: +!MESSAGE +!MESSAGE NMAKE /f "7z.mak" CFG="7z - Win32 Debug" +!MESSAGE +!MESSAGE Possible choices for configuration are: +!MESSAGE +!MESSAGE "7z - Win32 Release" (based on "Win32 (x86) Console Application") +!MESSAGE "7z - Win32 Debug" (based on "Win32 (x86) Console Application") +!MESSAGE + +# Begin Project +# PROP AllowPerConfigDependencies 0 +# PROP Scc_ProjName "" +# PROP Scc_LocalPath "" +CPP=cl.exe +RSC=rc.exe + +!IF "$(CFG)" == "7z - Win32 Release" + +# PROP BASE Use_MFC 0 +# PROP BASE Use_Debug_Libraries 0 +# PROP BASE Output_Dir "Release" +# PROP BASE Intermediate_Dir "Release" +# PROP BASE Target_Dir "" +# PROP Use_MFC 0 +# PROP Use_Debug_Libraries 0 +# PROP Output_Dir "Release" +# PROP Intermediate_Dir "Release" +# PROP Ignore_Export_Lib 0 +# PROP Target_Dir "" +# ADD BASE CPP /nologo /W3 /GX /O2 /D "WIN32" /D "NDEBUG" /D "_CONSOLE" /D "_MBCS" /YX /FD /c +# ADD CPP /nologo /MD /W4 /WX /GX /O2 /D "NDEBUG" /D "WIN32" /D "_CONSOLE" /D "_UNICODE" /D "UNICODE" /FAcs /Yu"Precomp.h" /FD /c +# ADD BASE RSC /l 0x419 /d "NDEBUG" +# ADD RSC /l 0x419 /d "NDEBUG" +BSC32=bscmake.exe +# ADD BASE BSC32 /nologo +# ADD BSC32 /nologo +LINK32=link.exe +# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /machine:I386 +# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /machine:I386 /out:"c:\util\7zDec.exe" /opt:NOWIN98 +# SUBTRACT LINK32 /pdb:none + +!ELSEIF "$(CFG)" == "7z - Win32 Debug" + +# PROP BASE Use_MFC 0 +# PROP BASE Use_Debug_Libraries 1 +# PROP BASE Output_Dir "Debug" +# PROP BASE Intermediate_Dir "Debug" +# PROP BASE Target_Dir "" +# PROP Use_MFC 0 +# PROP Use_Debug_Libraries 1 +# PROP Output_Dir "Debug" +# PROP Intermediate_Dir "Debug" +# PROP Ignore_Export_Lib 0 +# PROP Target_Dir "" +# ADD BASE CPP /nologo /W3 /Gm /GX /ZI /Od /D "WIN32" /D "_DEBUG" /D "_CONSOLE" /D "_MBCS" /YX /FD /GZ /c +# ADD CPP /nologo /W4 /WX /Gm /GX /ZI /Od /D "_DEBUG" /D "_SZ_ALLOC_DEBUG2" /D "_SZ_NO_INT_64_A" /D "WIN32" /D "_CONSOLE" /D "_UNICODE" /D "UNICODE" /Yu"Precomp.h" /FD /GZ /c +# ADD BASE RSC /l 0x419 /d "_DEBUG" +# ADD RSC /l 0x419 /d "_DEBUG" +BSC32=bscmake.exe +# ADD BASE BSC32 /nologo +# ADD BSC32 /nologo +LINK32=link.exe +# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /debug /machine:I386 /pdbtype:sept +# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /debug /machine:I386 /out:"c:\util\7zDec.exe" /pdbtype:sept + +!ENDIF + +# Begin Target + +# Name "7z - Win32 Release" +# Name "7z - Win32 Debug" +# Begin Group "Common" + +# PROP Default_Filter "" +# Begin Source File + +SOURCE=..\..\7z.h +# End Source File +# Begin Source File + +SOURCE=..\..\7zAlloc.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zAlloc.h +# End Source File +# Begin Source File + +SOURCE=..\..\7zArcIn.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zBuf.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zBuf.h +# End Source File +# Begin Source File + +SOURCE=..\..\7zCrc.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zCrc.h +# End Source File +# Begin Source File + +SOURCE=..\..\7zCrcOpt.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zDec.c +# ADD CPP /D "_7ZIP_PPMD_SUPPPORT" +# End Source File +# Begin Source File + +SOURCE=..\..\7zFile.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zFile.h +# End Source File +# Begin Source File + +SOURCE=..\..\7zStream.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zTypes.h +# End Source File +# Begin Source File + +SOURCE=..\..\Bcj2.c +# End Source File +# Begin Source File + +SOURCE=..\..\Bcj2.h +# End Source File +# Begin Source File + +SOURCE=..\..\Bra.c +# End Source File +# Begin Source File + +SOURCE=..\..\Bra.h +# End Source File +# Begin Source File + +SOURCE=..\..\Bra86.c +# End Source File +# Begin Source File + +SOURCE=..\..\BraIA64.c +# End Source File +# Begin Source File + +SOURCE=..\..\CpuArch.c +# End Source File +# Begin Source File + +SOURCE=..\..\CpuArch.h +# End Source File +# Begin Source File + +SOURCE=..\..\Delta.c +# End Source File +# Begin Source File + +SOURCE=..\..\Delta.h +# End Source File +# Begin Source File + +SOURCE=..\..\Lzma2Dec.c +# End Source File +# Begin Source File + +SOURCE=..\..\Lzma2Dec.h +# End Source File +# Begin Source File + +SOURCE=..\..\LzmaDec.c +# End Source File +# Begin Source File + +SOURCE=..\..\LzmaDec.h +# End Source File +# Begin Source File + +SOURCE=..\..\Ppmd.h +# End Source File +# Begin Source File + +SOURCE=..\..\Ppmd7.c +# End Source File +# Begin Source File + +SOURCE=..\..\Ppmd7.h +# End Source File +# Begin Source File + +SOURCE=..\..\Ppmd7Dec.c +# End Source File +# End Group +# Begin Group "Spec" + +# PROP Default_Filter "" +# Begin Source File + +SOURCE=..\..\Compiler.h +# End Source File +# Begin Source File + +SOURCE=.\Precomp.c +# ADD CPP /Yc"Precomp.h" +# End Source File +# Begin Source File + +SOURCE=.\Precomp.h +# End Source File +# End Group +# Begin Source File + +SOURCE=.\7zMain.c +# End Source File +# End Target +# End Project diff --git a/deps/LZMA-SDK/C/Util/7z/7z.dsw b/deps/LZMA-SDK/C/Util/7z/7z.dsw index 23089fb79..848d13cb1 100644 --- a/deps/LZMA-SDK/C/Util/7z/7z.dsw +++ b/deps/LZMA-SDK/C/Util/7z/7z.dsw @@ -1,29 +1,29 @@ -Microsoft Developer Studio Workspace File, Format Version 6.00 -# WARNING: DO NOT EDIT OR DELETE THIS WORKSPACE FILE! - -############################################################################### - -Project: "7z"=.\7z.dsp - Package Owner=<4> - -Package=<5> -{{{ -}}} - -Package=<4> -{{{ -}}} - -############################################################################### - -Global: - -Package=<5> -{{{ -}}} - -Package=<3> -{{{ -}}} - -############################################################################### - +Microsoft Developer Studio Workspace File, Format Version 6.00 +# WARNING: DO NOT EDIT OR DELETE THIS WORKSPACE FILE! + +############################################################################### + +Project: "7z"=.\7z.dsp - Package Owner=<4> + +Package=<5> +{{{ +}}} + +Package=<4> +{{{ +}}} + +############################################################################### + +Global: + +Package=<5> +{{{ +}}} + +Package=<3> +{{{ +}}} + +############################################################################### + diff --git a/deps/LZMA-SDK/C/Util/7z/7zMain.c b/deps/LZMA-SDK/C/Util/7z/7zMain.c index 3ab63a3d9..9d555098e 100644 --- a/deps/LZMA-SDK/C/Util/7z/7zMain.c +++ b/deps/LZMA-SDK/C/Util/7z/7zMain.c @@ -1,887 +1,887 @@ -/* 7zMain.c - Test application for 7z Decoder -2021-04-29 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include -#include - -#include "../../CpuArch.h" - -#include "../../7z.h" -#include "../../7zAlloc.h" -#include "../../7zBuf.h" -#include "../../7zCrc.h" -#include "../../7zFile.h" -#include "../../7zVersion.h" - -#ifndef USE_WINDOWS_FILE -/* for mkdir */ -#ifdef _WIN32 -#include -#else -#include -#include -#ifdef __GNUC__ -#include -#endif -#include -// #include -#include -#include -#endif -#endif - - -#define kInputBufSize ((size_t)1 << 18) - -static const ISzAlloc g_Alloc = { SzAlloc, SzFree }; - - -static void Print(const char *s) -{ - fputs(s, stdout); -} - - -static int Buf_EnsureSize(CBuf *dest, size_t size) -{ - if (dest->size >= size) - return 1; - Buf_Free(dest, &g_Alloc); - return Buf_Create(dest, size, &g_Alloc); -} - -#ifndef _WIN32 -#define _USE_UTF8 -#endif - -/* #define _USE_UTF8 */ - -#ifdef _USE_UTF8 - -#define _UTF8_START(n) (0x100 - (1 << (7 - (n)))) - -#define _UTF8_RANGE(n) (((UInt32)1) << ((n) * 5 + 6)) - -#define _UTF8_HEAD(n, val) ((Byte)(_UTF8_START(n) + (val >> (6 * (n))))) -#define _UTF8_CHAR(n, val) ((Byte)(0x80 + (((val) >> (6 * (n))) & 0x3F))) - -static size_t Utf16_To_Utf8_Calc(const UInt16 *src, const UInt16 *srcLim) -{ - size_t size = 0; - for (;;) - { - UInt32 val; - if (src == srcLim) - return size; - - size++; - val = *src++; - - if (val < 0x80) - continue; - - if (val < _UTF8_RANGE(1)) - { - size++; - continue; - } - - if (val >= 0xD800 && val < 0xDC00 && src != srcLim) - { - UInt32 c2 = *src; - if (c2 >= 0xDC00 && c2 < 0xE000) - { - src++; - size += 3; - continue; - } - } - - size += 2; - } -} - -static Byte *Utf16_To_Utf8(Byte *dest, const UInt16 *src, const UInt16 *srcLim) -{ - for (;;) - { - UInt32 val; - if (src == srcLim) - return dest; - - val = *src++; - - if (val < 0x80) - { - *dest++ = (Byte)val; - continue; - } - - if (val < _UTF8_RANGE(1)) - { - dest[0] = _UTF8_HEAD(1, val); - dest[1] = _UTF8_CHAR(0, val); - dest += 2; - continue; - } - - if (val >= 0xD800 && val < 0xDC00 && src != srcLim) - { - UInt32 c2 = *src; - if (c2 >= 0xDC00 && c2 < 0xE000) - { - src++; - val = (((val - 0xD800) << 10) | (c2 - 0xDC00)) + 0x10000; - dest[0] = _UTF8_HEAD(3, val); - dest[1] = _UTF8_CHAR(2, val); - dest[2] = _UTF8_CHAR(1, val); - dest[3] = _UTF8_CHAR(0, val); - dest += 4; - continue; - } - } - - dest[0] = _UTF8_HEAD(2, val); - dest[1] = _UTF8_CHAR(1, val); - dest[2] = _UTF8_CHAR(0, val); - dest += 3; - } -} - -static SRes Utf16_To_Utf8Buf(CBuf *dest, const UInt16 *src, size_t srcLen) -{ - size_t destLen = Utf16_To_Utf8_Calc(src, src + srcLen); - destLen += 1; - if (!Buf_EnsureSize(dest, destLen)) - return SZ_ERROR_MEM; - *Utf16_To_Utf8(dest->data, src, src + srcLen) = 0; - return SZ_OK; -} - -#endif - -static SRes Utf16_To_Char(CBuf *buf, const UInt16 *s - #ifndef _USE_UTF8 - , UINT codePage - #endif - ) -{ - unsigned len = 0; - for (len = 0; s[len] != 0; len++) {} - - #ifndef _USE_UTF8 - { - const unsigned size = len * 3 + 100; - if (!Buf_EnsureSize(buf, size)) - return SZ_ERROR_MEM; - { - buf->data[0] = 0; - if (len != 0) - { - const char defaultChar = '_'; - BOOL defUsed; - const unsigned numChars = (unsigned)WideCharToMultiByte( - codePage, 0, (LPCWSTR)s, (int)len, (char *)buf->data, (int)size, &defaultChar, &defUsed); - if (numChars == 0 || numChars >= size) - return SZ_ERROR_FAIL; - buf->data[numChars] = 0; - } - return SZ_OK; - } - } - #else - return Utf16_To_Utf8Buf(buf, s, len); - #endif -} - -#ifdef _WIN32 - #ifndef USE_WINDOWS_FILE - static UINT g_FileCodePage = CP_ACP; - #define MY_FILE_CODE_PAGE_PARAM ,g_FileCodePage - #endif -#else - #define MY_FILE_CODE_PAGE_PARAM -#endif - -static WRes MyCreateDir(const UInt16 *name) -{ - #ifdef USE_WINDOWS_FILE - - return CreateDirectoryW((LPCWSTR)name, NULL) ? 0 : GetLastError(); - - #else - - CBuf buf; - WRes res; - Buf_Init(&buf); - RINOK(Utf16_To_Char(&buf, name MY_FILE_CODE_PAGE_PARAM)); - - res = - #ifdef _WIN32 - _mkdir((const char *)buf.data) - #else - mkdir((const char *)buf.data, 0777) - #endif - == 0 ? 0 : errno; - Buf_Free(&buf, &g_Alloc); - return res; - - #endif -} - -static WRes OutFile_OpenUtf16(CSzFile *p, const UInt16 *name) -{ - #ifdef USE_WINDOWS_FILE - return OutFile_OpenW(p, (LPCWSTR)name); - #else - CBuf buf; - WRes res; - Buf_Init(&buf); - RINOK(Utf16_To_Char(&buf, name MY_FILE_CODE_PAGE_PARAM)); - res = OutFile_Open(p, (const char *)buf.data); - Buf_Free(&buf, &g_Alloc); - return res; - #endif -} - - -static SRes PrintString(const UInt16 *s) -{ - CBuf buf; - SRes res; - Buf_Init(&buf); - res = Utf16_To_Char(&buf, s - #ifndef _USE_UTF8 - , CP_OEMCP - #endif - ); - if (res == SZ_OK) - Print((const char *)buf.data); - Buf_Free(&buf, &g_Alloc); - return res; -} - -static void UInt64ToStr(UInt64 value, char *s, int numDigits) -{ - char temp[32]; - int pos = 0; - do - { - temp[pos++] = (char)('0' + (unsigned)(value % 10)); - value /= 10; - } - while (value != 0); - - for (numDigits -= pos; numDigits > 0; numDigits--) - *s++ = ' '; - - do - *s++ = temp[--pos]; - while (pos); - *s = '\0'; -} - -static char *UIntToStr(char *s, unsigned value, int numDigits) -{ - char temp[16]; - int pos = 0; - do - temp[pos++] = (char)('0' + (value % 10)); - while (value /= 10); - - for (numDigits -= pos; numDigits > 0; numDigits--) - *s++ = '0'; - - do - *s++ = temp[--pos]; - while (pos); - *s = '\0'; - return s; -} - -static void UIntToStr_2(char *s, unsigned value) -{ - s[0] = (char)('0' + (value / 10)); - s[1] = (char)('0' + (value % 10)); -} - - -#define PERIOD_4 (4 * 365 + 1) -#define PERIOD_100 (PERIOD_4 * 25 - 1) -#define PERIOD_400 (PERIOD_100 * 4 + 1) - - - -#ifndef _WIN32 - -// MS uses long for BOOL, but long is 32-bit in MS. So we use int. -// typedef long BOOL; -typedef int BOOL; - -typedef struct _FILETIME -{ - DWORD dwLowDateTime; - DWORD dwHighDateTime; -} FILETIME; - -static LONG TIME_GetBias() -{ - time_t utc = time(NULL); - struct tm *ptm = localtime(&utc); - int localdaylight = ptm->tm_isdst; /* daylight for local timezone */ - ptm = gmtime(&utc); - ptm->tm_isdst = localdaylight; /* use local daylight, not that of Greenwich */ - LONG bias = (int)(mktime(ptm)-utc); - return bias; -} - -#define TICKS_PER_SEC 10000000 - -#define GET_TIME_64(pft) ((pft)->dwLowDateTime | ((UInt64)(pft)->dwHighDateTime << 32)) - -#define SET_FILETIME(ft, v64) \ - (ft)->dwLowDateTime = (DWORD)v64; \ - (ft)->dwHighDateTime = (DWORD)(v64 >> 32); - -#define WINAPI -#define TRUE 1 - -static BOOL WINAPI FileTimeToLocalFileTime(const FILETIME *fileTime, FILETIME *localFileTime) -{ - UInt64 v = GET_TIME_64(fileTime); - v = (UInt64)((Int64)v - (Int64)TIME_GetBias() * TICKS_PER_SEC); - SET_FILETIME(localFileTime, v); - return TRUE; -} - -static const UInt32 kNumTimeQuantumsInSecond = 10000000; -static const UInt32 kFileTimeStartYear = 1601; -static const UInt32 kUnixTimeStartYear = 1970; -static const UInt64 kUnixTimeOffset = - (UInt64)60 * 60 * 24 * (89 + 365 * (kUnixTimeStartYear - kFileTimeStartYear)); - -static Int64 Time_FileTimeToUnixTime64(const FILETIME *ft) -{ - UInt64 winTime = GET_TIME_64(ft); - return (Int64)(winTime / kNumTimeQuantumsInSecond) - (Int64)kUnixTimeOffset; -} - -#if defined(_AIX) - #define MY_ST_TIMESPEC st_timespec -#else - #define MY_ST_TIMESPEC timespec -#endif - -static void FILETIME_To_timespec(const FILETIME *ft, struct MY_ST_TIMESPEC *ts) -{ - if (ft) - { - const Int64 sec = Time_FileTimeToUnixTime64(ft); - // time_t is long - const time_t sec2 = (time_t)sec; - if (sec2 == sec) - { - ts->tv_sec = sec2; - UInt64 winTime = GET_TIME_64(ft); - ts->tv_nsec = (long)((winTime % 10000000) * 100);; - return; - } - } - // else - { - ts->tv_sec = 0; - // ts.tv_nsec = UTIME_NOW; // set to the current time - ts->tv_nsec = UTIME_OMIT; // keep old timesptamp - } -} - -static WRes Set_File_FILETIME(const UInt16 *name, const FILETIME *mTime) -{ - struct timespec times[2]; - - const int flags = 0; // follow link - // = AT_SYMLINK_NOFOLLOW; // don't follow link - - CBuf buf; - int res; - Buf_Init(&buf); - RINOK(Utf16_To_Char(&buf, name MY_FILE_CODE_PAGE_PARAM)); - FILETIME_To_timespec(NULL, ×[0]); - FILETIME_To_timespec(mTime, ×[1]); - res = utimensat(AT_FDCWD, (const char *)buf.data, times, flags); - Buf_Free(&buf, &g_Alloc); - if (res == 0) - return 0; - return errno; -} - -#endif - -static void NtfsFileTime_to_FILETIME(const CNtfsFileTime *t, FILETIME *ft) -{ - ft->dwLowDateTime = (DWORD)(t->Low); - ft->dwHighDateTime = (DWORD)(t->High); -} - -static void ConvertFileTimeToString(const CNtfsFileTime *nTime, char *s) -{ - unsigned year, mon, hour, min, sec; - Byte ms[] = { 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 }; - unsigned t; - UInt32 v; - // UInt64 v64 = nt->Low | ((UInt64)nt->High << 32); - UInt64 v64; - { - FILETIME fileTime, locTime; - NtfsFileTime_to_FILETIME(nTime, &fileTime); - if (!FileTimeToLocalFileTime(&fileTime, &locTime)) - { - locTime.dwHighDateTime = - locTime.dwLowDateTime = 0; - } - v64 = locTime.dwLowDateTime | ((UInt64)locTime.dwHighDateTime << 32); - } - v64 /= 10000000; - sec = (unsigned)(v64 % 60); v64 /= 60; - min = (unsigned)(v64 % 60); v64 /= 60; - hour = (unsigned)(v64 % 24); v64 /= 24; - - v = (UInt32)v64; - - year = (unsigned)(1601 + v / PERIOD_400 * 400); - v %= PERIOD_400; - - t = v / PERIOD_100; if (t == 4) t = 3; year += t * 100; v -= t * PERIOD_100; - t = v / PERIOD_4; if (t == 25) t = 24; year += t * 4; v -= t * PERIOD_4; - t = v / 365; if (t == 4) t = 3; year += t; v -= t * 365; - - if (year % 4 == 0 && (year % 100 != 0 || year % 400 == 0)) - ms[1] = 29; - for (mon = 0;; mon++) - { - unsigned d = ms[mon]; - if (v < d) - break; - v -= d; - } - s = UIntToStr(s, year, 4); *s++ = '-'; - UIntToStr_2(s, mon + 1); s[2] = '-'; s += 3; - UIntToStr_2(s, (unsigned)v + 1); s[2] = ' '; s += 3; - UIntToStr_2(s, hour); s[2] = ':'; s += 3; - UIntToStr_2(s, min); s[2] = ':'; s += 3; - UIntToStr_2(s, sec); s[2] = 0; -} - -static void PrintLF() -{ - Print("\n"); -} - -static void PrintError(char *s) -{ - Print("\nERROR: "); - Print(s); - PrintLF(); -} - -static void PrintError_WRes(const char *message, WRes wres) -{ - Print("\nERROR: "); - Print(message); - PrintLF(); - { - char s[32]; - UIntToStr(s, (unsigned)wres, 1); - Print("System error code: "); - Print(s); - } - // sprintf(buffer + strlen(buffer), "\nSystem error code: %d", (unsigned)wres); - #ifdef _WIN32 - { - char *s = NULL; - if (FormatMessageA(FORMAT_MESSAGE_ALLOCATE_BUFFER | - FORMAT_MESSAGE_FROM_SYSTEM | FORMAT_MESSAGE_IGNORE_INSERTS, - NULL, wres, 0, (LPSTR) &s, 0, NULL) != 0 && s) - { - Print(" : "); - Print(s); - LocalFree(s); - } - } - #else - { - const char *s = strerror(wres); - if (s) - { - Print(" : "); - Print(s); - } - } - #endif - PrintLF(); -} - -static void GetAttribString(UInt32 wa, BoolInt isDir, char *s) -{ - #ifdef USE_WINDOWS_FILE - s[0] = (char)(((wa & FILE_ATTRIBUTE_DIRECTORY) != 0 || isDir) ? 'D' : '.'); - s[1] = (char)(((wa & FILE_ATTRIBUTE_READONLY ) != 0) ? 'R': '.'); - s[2] = (char)(((wa & FILE_ATTRIBUTE_HIDDEN ) != 0) ? 'H': '.'); - s[3] = (char)(((wa & FILE_ATTRIBUTE_SYSTEM ) != 0) ? 'S': '.'); - s[4] = (char)(((wa & FILE_ATTRIBUTE_ARCHIVE ) != 0) ? 'A': '.'); - s[5] = 0; - #else - s[0] = (char)(((wa & (1 << 4)) != 0 || isDir) ? 'D' : '.'); - s[1] = 0; - #endif -} - - -// #define NUM_PARENTS_MAX 128 - -int MY_CDECL main(int numargs, char *args[]) -{ - ISzAlloc allocImp; - ISzAlloc allocTempImp; - - CFileInStream archiveStream; - CLookToRead2 lookStream; - CSzArEx db; - SRes res; - UInt16 *temp = NULL; - size_t tempSize = 0; - // UInt32 parents[NUM_PARENTS_MAX]; - - Print("\n7z Decoder " MY_VERSION_CPU " : " MY_COPYRIGHT_DATE "\n\n"); - - if (numargs == 1) - { - Print( - "Usage: 7zDec \n\n" - "\n" - " e: Extract files from archive (without using directory names)\n" - " l: List contents of archive\n" - " t: Test integrity of archive\n" - " x: eXtract files with full paths\n"); - return 0; - } - - if (numargs < 3) - { - PrintError("incorrect command"); - return 1; - } - - #if defined(_WIN32) && !defined(USE_WINDOWS_FILE) && !defined(UNDER_CE) - g_FileCodePage = AreFileApisANSI() ? CP_ACP : CP_OEMCP; - #endif - - - allocImp = g_Alloc; - allocTempImp = g_Alloc; - - { - WRes wres = - #ifdef UNDER_CE - InFile_OpenW(&archiveStream.file, L"\test.7z"); // change it - #else - InFile_Open(&archiveStream.file, args[2]); - #endif - if (wres != 0) - { - PrintError_WRes("cannot open input file", wres); - return 1; - } - } - - FileInStream_CreateVTable(&archiveStream); - archiveStream.wres = 0; - LookToRead2_CreateVTable(&lookStream, False); - lookStream.buf = NULL; - - res = SZ_OK; - - { - lookStream.buf = (Byte *)ISzAlloc_Alloc(&allocImp, kInputBufSize); - if (!lookStream.buf) - res = SZ_ERROR_MEM; - else - { - lookStream.bufSize = kInputBufSize; - lookStream.realStream = &archiveStream.vt; - LookToRead2_Init(&lookStream); - } - } - - CrcGenerateTable(); - - SzArEx_Init(&db); - - if (res == SZ_OK) - { - res = SzArEx_Open(&db, &lookStream.vt, &allocImp, &allocTempImp); - } - - if (res == SZ_OK) - { - char *command = args[1]; - int listCommand = 0, testCommand = 0, fullPaths = 0; - - if (strcmp(command, "l") == 0) listCommand = 1; - else if (strcmp(command, "t") == 0) testCommand = 1; - else if (strcmp(command, "e") == 0) { } - else if (strcmp(command, "x") == 0) { fullPaths = 1; } - else - { - PrintError("incorrect command"); - res = SZ_ERROR_FAIL; - } - - if (res == SZ_OK) - { - UInt32 i; - - /* - if you need cache, use these 3 variables. - if you use external function, you can make these variable as static. - */ - UInt32 blockIndex = 0xFFFFFFFF; /* it can have any value before first call (if outBuffer = 0) */ - Byte *outBuffer = 0; /* it must be 0 before first call for each new archive. */ - size_t outBufferSize = 0; /* it can have any value before first call (if outBuffer = 0) */ - - for (i = 0; i < db.NumFiles; i++) - { - size_t offset = 0; - size_t outSizeProcessed = 0; - // const CSzFileItem *f = db.Files + i; - size_t len; - const BoolInt isDir = SzArEx_IsDir(&db, i); - if (listCommand == 0 && isDir && !fullPaths) - continue; - len = SzArEx_GetFileNameUtf16(&db, i, NULL); - // len = SzArEx_GetFullNameLen(&db, i); - - if (len > tempSize) - { - SzFree(NULL, temp); - tempSize = len; - temp = (UInt16 *)SzAlloc(NULL, tempSize * sizeof(temp[0])); - if (!temp) - { - res = SZ_ERROR_MEM; - break; - } - } - - SzArEx_GetFileNameUtf16(&db, i, temp); - /* - if (SzArEx_GetFullNameUtf16_Back(&db, i, temp + len) != temp) - { - res = SZ_ERROR_FAIL; - break; - } - */ - - if (listCommand) - { - char attr[8], s[32], t[32]; - UInt64 fileSize; - - GetAttribString(SzBitWithVals_Check(&db.Attribs, i) ? db.Attribs.Vals[i] : 0, isDir, attr); - - fileSize = SzArEx_GetFileSize(&db, i); - UInt64ToStr(fileSize, s, 10); - - if (SzBitWithVals_Check(&db.MTime, i)) - ConvertFileTimeToString(&db.MTime.Vals[i], t); - else - { - size_t j; - for (j = 0; j < 19; j++) - t[j] = ' '; - t[j] = '\0'; - } - - Print(t); - Print(" "); - Print(attr); - Print(" "); - Print(s); - Print(" "); - res = PrintString(temp); - if (res != SZ_OK) - break; - if (isDir) - Print("/"); - PrintLF(); - continue; - } - - Print(testCommand ? - "T ": - "- "); - res = PrintString(temp); - if (res != SZ_OK) - break; - - if (isDir) - Print("/"); - else - { - res = SzArEx_Extract(&db, &lookStream.vt, i, - &blockIndex, &outBuffer, &outBufferSize, - &offset, &outSizeProcessed, - &allocImp, &allocTempImp); - if (res != SZ_OK) - break; - } - - if (!testCommand) - { - CSzFile outFile; - size_t processedSize; - size_t j; - UInt16 *name = (UInt16 *)temp; - const UInt16 *destPath = (const UInt16 *)name; - - for (j = 0; name[j] != 0; j++) - if (name[j] == '/') - { - if (fullPaths) - { - name[j] = 0; - MyCreateDir(name); - name[j] = CHAR_PATH_SEPARATOR; - } - else - destPath = name + j + 1; - } - - if (isDir) - { - MyCreateDir(destPath); - PrintLF(); - continue; - } - else - { - WRes wres = OutFile_OpenUtf16(&outFile, destPath); - if (wres != 0) - { - PrintError_WRes("cannot open output file", wres); - res = SZ_ERROR_FAIL; - break; - } - } - - processedSize = outSizeProcessed; - - { - WRes wres = File_Write(&outFile, outBuffer + offset, &processedSize); - if (wres != 0 || processedSize != outSizeProcessed) - { - PrintError_WRes("cannot write output file", wres); - res = SZ_ERROR_FAIL; - break; - } - } - - { - FILETIME mtime; - FILETIME *mtimePtr = NULL; - - #ifdef USE_WINDOWS_FILE - FILETIME ctime; - FILETIME *ctimePtr = NULL; - #endif - - if (SzBitWithVals_Check(&db.MTime, i)) - { - const CNtfsFileTime *t = &db.MTime.Vals[i]; - mtime.dwLowDateTime = (DWORD)(t->Low); - mtime.dwHighDateTime = (DWORD)(t->High); - mtimePtr = &mtime; - } - - #ifdef USE_WINDOWS_FILE - if (SzBitWithVals_Check(&db.CTime, i)) - { - const CNtfsFileTime *t = &db.CTime.Vals[i]; - ctime.dwLowDateTime = (DWORD)(t->Low); - ctime.dwHighDateTime = (DWORD)(t->High); - ctimePtr = &ctime; - } - - if (mtimePtr || ctimePtr) - SetFileTime(outFile.handle, ctimePtr, NULL, mtimePtr); - #endif - - { - WRes wres = File_Close(&outFile); - if (wres != 0) - { - PrintError_WRes("cannot close output file", wres); - res = SZ_ERROR_FAIL; - break; - } - } - - #ifndef USE_WINDOWS_FILE - #ifdef _WIN32 - mtimePtr = mtimePtr; - #else - if (mtimePtr) - Set_File_FILETIME(destPath, mtimePtr); - #endif - #endif - } - - #ifdef USE_WINDOWS_FILE - if (SzBitWithVals_Check(&db.Attribs, i)) - { - UInt32 attrib = db.Attribs.Vals[i]; - /* p7zip stores posix attributes in high 16 bits and adds 0x8000 as marker. - We remove posix bits, if we detect posix mode field */ - if ((attrib & 0xF0000000) != 0) - attrib &= 0x7FFF; - SetFileAttributesW((LPCWSTR)destPath, attrib); - } - #endif - } - PrintLF(); - } - ISzAlloc_Free(&allocImp, outBuffer); - } - } - - SzFree(NULL, temp); - SzArEx_Free(&db, &allocImp); - ISzAlloc_Free(&allocImp, lookStream.buf); - - File_Close(&archiveStream.file); - - if (res == SZ_OK) - { - Print("\nEverything is Ok\n"); - return 0; - } - - if (res == SZ_ERROR_UNSUPPORTED) - PrintError("decoder doesn't support this archive"); - else if (res == SZ_ERROR_MEM) - PrintError("cannot allocate memory"); - else if (res == SZ_ERROR_CRC) - PrintError("CRC error"); - else if (res == SZ_ERROR_READ /* || archiveStream.Res != 0 */) - PrintError_WRes("Read Error", archiveStream.wres); - else - { - char s[32]; - UInt64ToStr((unsigned)res, s, 0); - PrintError(s); - } - - return 1; -} +/* 7zMain.c - Test application for 7z Decoder +2021-04-29 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include +#include + +#include "../../CpuArch.h" + +#include "../../7z.h" +#include "../../7zAlloc.h" +#include "../../7zBuf.h" +#include "../../7zCrc.h" +#include "../../7zFile.h" +#include "../../7zVersion.h" + +#ifndef USE_WINDOWS_FILE +/* for mkdir */ +#ifdef _WIN32 +#include +#else +#include +#include +#ifdef __GNUC__ +#include +#endif +#include +// #include +#include +#include +#endif +#endif + + +#define kInputBufSize ((size_t)1 << 18) + +static const ISzAlloc g_Alloc = { SzAlloc, SzFree }; + + +static void Print(const char *s) +{ + fputs(s, stdout); +} + + +static int Buf_EnsureSize(CBuf *dest, size_t size) +{ + if (dest->size >= size) + return 1; + Buf_Free(dest, &g_Alloc); + return Buf_Create(dest, size, &g_Alloc); +} + +#ifndef _WIN32 +#define _USE_UTF8 +#endif + +/* #define _USE_UTF8 */ + +#ifdef _USE_UTF8 + +#define _UTF8_START(n) (0x100 - (1 << (7 - (n)))) + +#define _UTF8_RANGE(n) (((UInt32)1) << ((n) * 5 + 6)) + +#define _UTF8_HEAD(n, val) ((Byte)(_UTF8_START(n) + (val >> (6 * (n))))) +#define _UTF8_CHAR(n, val) ((Byte)(0x80 + (((val) >> (6 * (n))) & 0x3F))) + +static size_t Utf16_To_Utf8_Calc(const UInt16 *src, const UInt16 *srcLim) +{ + size_t size = 0; + for (;;) + { + UInt32 val; + if (src == srcLim) + return size; + + size++; + val = *src++; + + if (val < 0x80) + continue; + + if (val < _UTF8_RANGE(1)) + { + size++; + continue; + } + + if (val >= 0xD800 && val < 0xDC00 && src != srcLim) + { + UInt32 c2 = *src; + if (c2 >= 0xDC00 && c2 < 0xE000) + { + src++; + size += 3; + continue; + } + } + + size += 2; + } +} + +static Byte *Utf16_To_Utf8(Byte *dest, const UInt16 *src, const UInt16 *srcLim) +{ + for (;;) + { + UInt32 val; + if (src == srcLim) + return dest; + + val = *src++; + + if (val < 0x80) + { + *dest++ = (Byte)val; + continue; + } + + if (val < _UTF8_RANGE(1)) + { + dest[0] = _UTF8_HEAD(1, val); + dest[1] = _UTF8_CHAR(0, val); + dest += 2; + continue; + } + + if (val >= 0xD800 && val < 0xDC00 && src != srcLim) + { + UInt32 c2 = *src; + if (c2 >= 0xDC00 && c2 < 0xE000) + { + src++; + val = (((val - 0xD800) << 10) | (c2 - 0xDC00)) + 0x10000; + dest[0] = _UTF8_HEAD(3, val); + dest[1] = _UTF8_CHAR(2, val); + dest[2] = _UTF8_CHAR(1, val); + dest[3] = _UTF8_CHAR(0, val); + dest += 4; + continue; + } + } + + dest[0] = _UTF8_HEAD(2, val); + dest[1] = _UTF8_CHAR(1, val); + dest[2] = _UTF8_CHAR(0, val); + dest += 3; + } +} + +static SRes Utf16_To_Utf8Buf(CBuf *dest, const UInt16 *src, size_t srcLen) +{ + size_t destLen = Utf16_To_Utf8_Calc(src, src + srcLen); + destLen += 1; + if (!Buf_EnsureSize(dest, destLen)) + return SZ_ERROR_MEM; + *Utf16_To_Utf8(dest->data, src, src + srcLen) = 0; + return SZ_OK; +} + +#endif + +static SRes Utf16_To_Char(CBuf *buf, const UInt16 *s + #ifndef _USE_UTF8 + , UINT codePage + #endif + ) +{ + unsigned len = 0; + for (len = 0; s[len] != 0; len++) {} + + #ifndef _USE_UTF8 + { + const unsigned size = len * 3 + 100; + if (!Buf_EnsureSize(buf, size)) + return SZ_ERROR_MEM; + { + buf->data[0] = 0; + if (len != 0) + { + const char defaultChar = '_'; + BOOL defUsed; + const unsigned numChars = (unsigned)WideCharToMultiByte( + codePage, 0, (LPCWSTR)s, (int)len, (char *)buf->data, (int)size, &defaultChar, &defUsed); + if (numChars == 0 || numChars >= size) + return SZ_ERROR_FAIL; + buf->data[numChars] = 0; + } + return SZ_OK; + } + } + #else + return Utf16_To_Utf8Buf(buf, s, len); + #endif +} + +#ifdef _WIN32 + #ifndef USE_WINDOWS_FILE + static UINT g_FileCodePage = CP_ACP; + #define MY_FILE_CODE_PAGE_PARAM ,g_FileCodePage + #endif +#else + #define MY_FILE_CODE_PAGE_PARAM +#endif + +static WRes MyCreateDir(const UInt16 *name) +{ + #ifdef USE_WINDOWS_FILE + + return CreateDirectoryW((LPCWSTR)name, NULL) ? 0 : GetLastError(); + + #else + + CBuf buf; + WRes res; + Buf_Init(&buf); + RINOK(Utf16_To_Char(&buf, name MY_FILE_CODE_PAGE_PARAM)); + + res = + #ifdef _WIN32 + _mkdir((const char *)buf.data) + #else + mkdir((const char *)buf.data, 0777) + #endif + == 0 ? 0 : errno; + Buf_Free(&buf, &g_Alloc); + return res; + + #endif +} + +static WRes OutFile_OpenUtf16(CSzFile *p, const UInt16 *name) +{ + #ifdef USE_WINDOWS_FILE + return OutFile_OpenW(p, (LPCWSTR)name); + #else + CBuf buf; + WRes res; + Buf_Init(&buf); + RINOK(Utf16_To_Char(&buf, name MY_FILE_CODE_PAGE_PARAM)); + res = OutFile_Open(p, (const char *)buf.data); + Buf_Free(&buf, &g_Alloc); + return res; + #endif +} + + +static SRes PrintString(const UInt16 *s) +{ + CBuf buf; + SRes res; + Buf_Init(&buf); + res = Utf16_To_Char(&buf, s + #ifndef _USE_UTF8 + , CP_OEMCP + #endif + ); + if (res == SZ_OK) + Print((const char *)buf.data); + Buf_Free(&buf, &g_Alloc); + return res; +} + +static void UInt64ToStr(UInt64 value, char *s, int numDigits) +{ + char temp[32]; + int pos = 0; + do + { + temp[pos++] = (char)('0' + (unsigned)(value % 10)); + value /= 10; + } + while (value != 0); + + for (numDigits -= pos; numDigits > 0; numDigits--) + *s++ = ' '; + + do + *s++ = temp[--pos]; + while (pos); + *s = '\0'; +} + +static char *UIntToStr(char *s, unsigned value, int numDigits) +{ + char temp[16]; + int pos = 0; + do + temp[pos++] = (char)('0' + (value % 10)); + while (value /= 10); + + for (numDigits -= pos; numDigits > 0; numDigits--) + *s++ = '0'; + + do + *s++ = temp[--pos]; + while (pos); + *s = '\0'; + return s; +} + +static void UIntToStr_2(char *s, unsigned value) +{ + s[0] = (char)('0' + (value / 10)); + s[1] = (char)('0' + (value % 10)); +} + + +#define PERIOD_4 (4 * 365 + 1) +#define PERIOD_100 (PERIOD_4 * 25 - 1) +#define PERIOD_400 (PERIOD_100 * 4 + 1) + + + +#ifndef _WIN32 + +// MS uses long for BOOL, but long is 32-bit in MS. So we use int. +// typedef long BOOL; +typedef int BOOL; + +typedef struct _FILETIME +{ + DWORD dwLowDateTime; + DWORD dwHighDateTime; +} FILETIME; + +static LONG TIME_GetBias() +{ + time_t utc = time(NULL); + struct tm *ptm = localtime(&utc); + int localdaylight = ptm->tm_isdst; /* daylight for local timezone */ + ptm = gmtime(&utc); + ptm->tm_isdst = localdaylight; /* use local daylight, not that of Greenwich */ + LONG bias = (int)(mktime(ptm)-utc); + return bias; +} + +#define TICKS_PER_SEC 10000000 + +#define GET_TIME_64(pft) ((pft)->dwLowDateTime | ((UInt64)(pft)->dwHighDateTime << 32)) + +#define SET_FILETIME(ft, v64) \ + (ft)->dwLowDateTime = (DWORD)v64; \ + (ft)->dwHighDateTime = (DWORD)(v64 >> 32); + +#define WINAPI +#define TRUE 1 + +static BOOL WINAPI FileTimeToLocalFileTime(const FILETIME *fileTime, FILETIME *localFileTime) +{ + UInt64 v = GET_TIME_64(fileTime); + v = (UInt64)((Int64)v - (Int64)TIME_GetBias() * TICKS_PER_SEC); + SET_FILETIME(localFileTime, v); + return TRUE; +} + +static const UInt32 kNumTimeQuantumsInSecond = 10000000; +static const UInt32 kFileTimeStartYear = 1601; +static const UInt32 kUnixTimeStartYear = 1970; +static const UInt64 kUnixTimeOffset = + (UInt64)60 * 60 * 24 * (89 + 365 * (kUnixTimeStartYear - kFileTimeStartYear)); + +static Int64 Time_FileTimeToUnixTime64(const FILETIME *ft) +{ + UInt64 winTime = GET_TIME_64(ft); + return (Int64)(winTime / kNumTimeQuantumsInSecond) - (Int64)kUnixTimeOffset; +} + +#if defined(_AIX) + #define MY_ST_TIMESPEC st_timespec +#else + #define MY_ST_TIMESPEC timespec +#endif + +static void FILETIME_To_timespec(const FILETIME *ft, struct MY_ST_TIMESPEC *ts) +{ + if (ft) + { + const Int64 sec = Time_FileTimeToUnixTime64(ft); + // time_t is long + const time_t sec2 = (time_t)sec; + if (sec2 == sec) + { + ts->tv_sec = sec2; + UInt64 winTime = GET_TIME_64(ft); + ts->tv_nsec = (long)((winTime % 10000000) * 100);; + return; + } + } + // else + { + ts->tv_sec = 0; + // ts.tv_nsec = UTIME_NOW; // set to the current time + ts->tv_nsec = UTIME_OMIT; // keep old timesptamp + } +} + +static WRes Set_File_FILETIME(const UInt16 *name, const FILETIME *mTime) +{ + struct timespec times[2]; + + const int flags = 0; // follow link + // = AT_SYMLINK_NOFOLLOW; // don't follow link + + CBuf buf; + int res; + Buf_Init(&buf); + RINOK(Utf16_To_Char(&buf, name MY_FILE_CODE_PAGE_PARAM)); + FILETIME_To_timespec(NULL, ×[0]); + FILETIME_To_timespec(mTime, ×[1]); + res = utimensat(AT_FDCWD, (const char *)buf.data, times, flags); + Buf_Free(&buf, &g_Alloc); + if (res == 0) + return 0; + return errno; +} + +#endif + +static void NtfsFileTime_to_FILETIME(const CNtfsFileTime *t, FILETIME *ft) +{ + ft->dwLowDateTime = (DWORD)(t->Low); + ft->dwHighDateTime = (DWORD)(t->High); +} + +static void ConvertFileTimeToString(const CNtfsFileTime *nTime, char *s) +{ + unsigned year, mon, hour, min, sec; + Byte ms[] = { 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 }; + unsigned t; + UInt32 v; + // UInt64 v64 = nt->Low | ((UInt64)nt->High << 32); + UInt64 v64; + { + FILETIME fileTime, locTime; + NtfsFileTime_to_FILETIME(nTime, &fileTime); + if (!FileTimeToLocalFileTime(&fileTime, &locTime)) + { + locTime.dwHighDateTime = + locTime.dwLowDateTime = 0; + } + v64 = locTime.dwLowDateTime | ((UInt64)locTime.dwHighDateTime << 32); + } + v64 /= 10000000; + sec = (unsigned)(v64 % 60); v64 /= 60; + min = (unsigned)(v64 % 60); v64 /= 60; + hour = (unsigned)(v64 % 24); v64 /= 24; + + v = (UInt32)v64; + + year = (unsigned)(1601 + v / PERIOD_400 * 400); + v %= PERIOD_400; + + t = v / PERIOD_100; if (t == 4) t = 3; year += t * 100; v -= t * PERIOD_100; + t = v / PERIOD_4; if (t == 25) t = 24; year += t * 4; v -= t * PERIOD_4; + t = v / 365; if (t == 4) t = 3; year += t; v -= t * 365; + + if (year % 4 == 0 && (year % 100 != 0 || year % 400 == 0)) + ms[1] = 29; + for (mon = 0;; mon++) + { + unsigned d = ms[mon]; + if (v < d) + break; + v -= d; + } + s = UIntToStr(s, year, 4); *s++ = '-'; + UIntToStr_2(s, mon + 1); s[2] = '-'; s += 3; + UIntToStr_2(s, (unsigned)v + 1); s[2] = ' '; s += 3; + UIntToStr_2(s, hour); s[2] = ':'; s += 3; + UIntToStr_2(s, min); s[2] = ':'; s += 3; + UIntToStr_2(s, sec); s[2] = 0; +} + +static void PrintLF() +{ + Print("\n"); +} + +static void PrintError(char *s) +{ + Print("\nERROR: "); + Print(s); + PrintLF(); +} + +static void PrintError_WRes(const char *message, WRes wres) +{ + Print("\nERROR: "); + Print(message); + PrintLF(); + { + char s[32]; + UIntToStr(s, (unsigned)wres, 1); + Print("System error code: "); + Print(s); + } + // sprintf(buffer + strlen(buffer), "\nSystem error code: %d", (unsigned)wres); + #ifdef _WIN32 + { + char *s = NULL; + if (FormatMessageA(FORMAT_MESSAGE_ALLOCATE_BUFFER | + FORMAT_MESSAGE_FROM_SYSTEM | FORMAT_MESSAGE_IGNORE_INSERTS, + NULL, wres, 0, (LPSTR) &s, 0, NULL) != 0 && s) + { + Print(" : "); + Print(s); + LocalFree(s); + } + } + #else + { + const char *s = strerror(wres); + if (s) + { + Print(" : "); + Print(s); + } + } + #endif + PrintLF(); +} + +static void GetAttribString(UInt32 wa, BoolInt isDir, char *s) +{ + #ifdef USE_WINDOWS_FILE + s[0] = (char)(((wa & FILE_ATTRIBUTE_DIRECTORY) != 0 || isDir) ? 'D' : '.'); + s[1] = (char)(((wa & FILE_ATTRIBUTE_READONLY ) != 0) ? 'R': '.'); + s[2] = (char)(((wa & FILE_ATTRIBUTE_HIDDEN ) != 0) ? 'H': '.'); + s[3] = (char)(((wa & FILE_ATTRIBUTE_SYSTEM ) != 0) ? 'S': '.'); + s[4] = (char)(((wa & FILE_ATTRIBUTE_ARCHIVE ) != 0) ? 'A': '.'); + s[5] = 0; + #else + s[0] = (char)(((wa & (1 << 4)) != 0 || isDir) ? 'D' : '.'); + s[1] = 0; + #endif +} + + +// #define NUM_PARENTS_MAX 128 + +int MY_CDECL main(int numargs, char *args[]) +{ + ISzAlloc allocImp; + ISzAlloc allocTempImp; + + CFileInStream archiveStream; + CLookToRead2 lookStream; + CSzArEx db; + SRes res; + UInt16 *temp = NULL; + size_t tempSize = 0; + // UInt32 parents[NUM_PARENTS_MAX]; + + Print("\n7z Decoder " MY_VERSION_CPU " : " MY_COPYRIGHT_DATE "\n\n"); + + if (numargs == 1) + { + Print( + "Usage: 7zDec \n\n" + "\n" + " e: Extract files from archive (without using directory names)\n" + " l: List contents of archive\n" + " t: Test integrity of archive\n" + " x: eXtract files with full paths\n"); + return 0; + } + + if (numargs < 3) + { + PrintError("incorrect command"); + return 1; + } + + #if defined(_WIN32) && !defined(USE_WINDOWS_FILE) && !defined(UNDER_CE) + g_FileCodePage = AreFileApisANSI() ? CP_ACP : CP_OEMCP; + #endif + + + allocImp = g_Alloc; + allocTempImp = g_Alloc; + + { + WRes wres = + #ifdef UNDER_CE + InFile_OpenW(&archiveStream.file, L"\test.7z"); // change it + #else + InFile_Open(&archiveStream.file, args[2]); + #endif + if (wres != 0) + { + PrintError_WRes("cannot open input file", wres); + return 1; + } + } + + FileInStream_CreateVTable(&archiveStream); + archiveStream.wres = 0; + LookToRead2_CreateVTable(&lookStream, False); + lookStream.buf = NULL; + + res = SZ_OK; + + { + lookStream.buf = (Byte *)ISzAlloc_Alloc(&allocImp, kInputBufSize); + if (!lookStream.buf) + res = SZ_ERROR_MEM; + else + { + lookStream.bufSize = kInputBufSize; + lookStream.realStream = &archiveStream.vt; + LookToRead2_Init(&lookStream); + } + } + + CrcGenerateTable(); + + SzArEx_Init(&db); + + if (res == SZ_OK) + { + res = SzArEx_Open(&db, &lookStream.vt, &allocImp, &allocTempImp); + } + + if (res == SZ_OK) + { + char *command = args[1]; + int listCommand = 0, testCommand = 0, fullPaths = 0; + + if (strcmp(command, "l") == 0) listCommand = 1; + else if (strcmp(command, "t") == 0) testCommand = 1; + else if (strcmp(command, "e") == 0) { } + else if (strcmp(command, "x") == 0) { fullPaths = 1; } + else + { + PrintError("incorrect command"); + res = SZ_ERROR_FAIL; + } + + if (res == SZ_OK) + { + UInt32 i; + + /* + if you need cache, use these 3 variables. + if you use external function, you can make these variable as static. + */ + UInt32 blockIndex = 0xFFFFFFFF; /* it can have any value before first call (if outBuffer = 0) */ + Byte *outBuffer = 0; /* it must be 0 before first call for each new archive. */ + size_t outBufferSize = 0; /* it can have any value before first call (if outBuffer = 0) */ + + for (i = 0; i < db.NumFiles; i++) + { + size_t offset = 0; + size_t outSizeProcessed = 0; + // const CSzFileItem *f = db.Files + i; + size_t len; + const BoolInt isDir = SzArEx_IsDir(&db, i); + if (listCommand == 0 && isDir && !fullPaths) + continue; + len = SzArEx_GetFileNameUtf16(&db, i, NULL); + // len = SzArEx_GetFullNameLen(&db, i); + + if (len > tempSize) + { + SzFree(NULL, temp); + tempSize = len; + temp = (UInt16 *)SzAlloc(NULL, tempSize * sizeof(temp[0])); + if (!temp) + { + res = SZ_ERROR_MEM; + break; + } + } + + SzArEx_GetFileNameUtf16(&db, i, temp); + /* + if (SzArEx_GetFullNameUtf16_Back(&db, i, temp + len) != temp) + { + res = SZ_ERROR_FAIL; + break; + } + */ + + if (listCommand) + { + char attr[8], s[32], t[32]; + UInt64 fileSize; + + GetAttribString(SzBitWithVals_Check(&db.Attribs, i) ? db.Attribs.Vals[i] : 0, isDir, attr); + + fileSize = SzArEx_GetFileSize(&db, i); + UInt64ToStr(fileSize, s, 10); + + if (SzBitWithVals_Check(&db.MTime, i)) + ConvertFileTimeToString(&db.MTime.Vals[i], t); + else + { + size_t j; + for (j = 0; j < 19; j++) + t[j] = ' '; + t[j] = '\0'; + } + + Print(t); + Print(" "); + Print(attr); + Print(" "); + Print(s); + Print(" "); + res = PrintString(temp); + if (res != SZ_OK) + break; + if (isDir) + Print("/"); + PrintLF(); + continue; + } + + Print(testCommand ? + "T ": + "- "); + res = PrintString(temp); + if (res != SZ_OK) + break; + + if (isDir) + Print("/"); + else + { + res = SzArEx_Extract(&db, &lookStream.vt, i, + &blockIndex, &outBuffer, &outBufferSize, + &offset, &outSizeProcessed, + &allocImp, &allocTempImp); + if (res != SZ_OK) + break; + } + + if (!testCommand) + { + CSzFile outFile; + size_t processedSize; + size_t j; + UInt16 *name = (UInt16 *)temp; + const UInt16 *destPath = (const UInt16 *)name; + + for (j = 0; name[j] != 0; j++) + if (name[j] == '/') + { + if (fullPaths) + { + name[j] = 0; + MyCreateDir(name); + name[j] = CHAR_PATH_SEPARATOR; + } + else + destPath = name + j + 1; + } + + if (isDir) + { + MyCreateDir(destPath); + PrintLF(); + continue; + } + else + { + WRes wres = OutFile_OpenUtf16(&outFile, destPath); + if (wres != 0) + { + PrintError_WRes("cannot open output file", wres); + res = SZ_ERROR_FAIL; + break; + } + } + + processedSize = outSizeProcessed; + + { + WRes wres = File_Write(&outFile, outBuffer + offset, &processedSize); + if (wres != 0 || processedSize != outSizeProcessed) + { + PrintError_WRes("cannot write output file", wres); + res = SZ_ERROR_FAIL; + break; + } + } + + { + FILETIME mtime; + FILETIME *mtimePtr = NULL; + + #ifdef USE_WINDOWS_FILE + FILETIME ctime; + FILETIME *ctimePtr = NULL; + #endif + + if (SzBitWithVals_Check(&db.MTime, i)) + { + const CNtfsFileTime *t = &db.MTime.Vals[i]; + mtime.dwLowDateTime = (DWORD)(t->Low); + mtime.dwHighDateTime = (DWORD)(t->High); + mtimePtr = &mtime; + } + + #ifdef USE_WINDOWS_FILE + if (SzBitWithVals_Check(&db.CTime, i)) + { + const CNtfsFileTime *t = &db.CTime.Vals[i]; + ctime.dwLowDateTime = (DWORD)(t->Low); + ctime.dwHighDateTime = (DWORD)(t->High); + ctimePtr = &ctime; + } + + if (mtimePtr || ctimePtr) + SetFileTime(outFile.handle, ctimePtr, NULL, mtimePtr); + #endif + + { + WRes wres = File_Close(&outFile); + if (wres != 0) + { + PrintError_WRes("cannot close output file", wres); + res = SZ_ERROR_FAIL; + break; + } + } + + #ifndef USE_WINDOWS_FILE + #ifdef _WIN32 + mtimePtr = mtimePtr; + #else + if (mtimePtr) + Set_File_FILETIME(destPath, mtimePtr); + #endif + #endif + } + + #ifdef USE_WINDOWS_FILE + if (SzBitWithVals_Check(&db.Attribs, i)) + { + UInt32 attrib = db.Attribs.Vals[i]; + /* p7zip stores posix attributes in high 16 bits and adds 0x8000 as marker. + We remove posix bits, if we detect posix mode field */ + if ((attrib & 0xF0000000) != 0) + attrib &= 0x7FFF; + SetFileAttributesW((LPCWSTR)destPath, attrib); + } + #endif + } + PrintLF(); + } + ISzAlloc_Free(&allocImp, outBuffer); + } + } + + SzFree(NULL, temp); + SzArEx_Free(&db, &allocImp); + ISzAlloc_Free(&allocImp, lookStream.buf); + + File_Close(&archiveStream.file); + + if (res == SZ_OK) + { + Print("\nEverything is Ok\n"); + return 0; + } + + if (res == SZ_ERROR_UNSUPPORTED) + PrintError("decoder doesn't support this archive"); + else if (res == SZ_ERROR_MEM) + PrintError("cannot allocate memory"); + else if (res == SZ_ERROR_CRC) + PrintError("CRC error"); + else if (res == SZ_ERROR_READ /* || archiveStream.Res != 0 */) + PrintError_WRes("Read Error", archiveStream.wres); + else + { + char s[32]; + UInt64ToStr((unsigned)res, s, 0); + PrintError(s); + } + + return 1; +} diff --git a/deps/LZMA-SDK/C/Util/7z/Precomp.c b/deps/LZMA-SDK/C/Util/7z/Precomp.c index 34b60f8fc..01605e3c2 100644 --- a/deps/LZMA-SDK/C/Util/7z/Precomp.c +++ b/deps/LZMA-SDK/C/Util/7z/Precomp.c @@ -1,4 +1,4 @@ -/* Precomp.c -- StdAfx -2013-01-21 : Igor Pavlov : Public domain */ - -#include "Precomp.h" +/* Precomp.c -- StdAfx +2013-01-21 : Igor Pavlov : Public domain */ + +#include "Precomp.h" diff --git a/deps/LZMA-SDK/C/Util/7z/Precomp.h b/deps/LZMA-SDK/C/Util/7z/Precomp.h index 9f398d08f..588a66f7e 100644 --- a/deps/LZMA-SDK/C/Util/7z/Precomp.h +++ b/deps/LZMA-SDK/C/Util/7z/Precomp.h @@ -1,10 +1,10 @@ -/* Precomp.h -- StdAfx -2013-06-16 : Igor Pavlov : Public domain */ - -#ifndef __7Z_PRECOMP_H -#define __7Z_PRECOMP_H - -#include "../../Compiler.h" -#include "../../7zTypes.h" - -#endif +/* Precomp.h -- StdAfx +2013-06-16 : Igor Pavlov : Public domain */ + +#ifndef __7Z_PRECOMP_H +#define __7Z_PRECOMP_H + +#include "../../Compiler.h" +#include "../../7zTypes.h" + +#endif diff --git a/deps/LZMA-SDK/C/Util/7z/makefile b/deps/LZMA-SDK/C/Util/7z/makefile index f4a54af73..9a49fd515 100644 --- a/deps/LZMA-SDK/C/Util/7z/makefile +++ b/deps/LZMA-SDK/C/Util/7z/makefile @@ -1,40 +1,40 @@ -CFLAGS = $(CFLAGS) -D_7ZIP_PPMD_SUPPPORT - -PROG = 7zDec.exe - -C_OBJS = \ - $O\7zAlloc.obj \ - $O\7zBuf.obj \ - $O\7zCrc.obj \ - $O\7zCrcOpt.obj \ - $O\7zFile.obj \ - $O\7zDec.obj \ - $O\7zArcIn.obj \ - $O\7zStream.obj \ - $O\Bcj2.obj \ - $O\Bra.obj \ - $O\Bra86.obj \ - $O\BraIA64.obj \ - $O\CpuArch.obj \ - $O\Delta.obj \ - $O\Lzma2Dec.obj \ - $O\LzmaDec.obj \ - $O\Ppmd7.obj \ - $O\Ppmd7Dec.obj \ - -7Z_OBJS = \ - $O\7zMain.obj \ - -OBJS = \ - $O\Precomp.obj \ - $(7Z_OBJS) \ - $(C_OBJS) \ - -!include "../../../CPP/Build.mak" - -$(7Z_OBJS): $(*B).c - $(CCOMPL_USE) -$(C_OBJS): ../../$(*B).c - $(CCOMPL_USE) -$O\Precomp.obj: Precomp.c - $(CCOMPL_PCH) +CFLAGS = $(CFLAGS) -D_7ZIP_PPMD_SUPPPORT + +PROG = 7zDec.exe + +C_OBJS = \ + $O\7zAlloc.obj \ + $O\7zBuf.obj \ + $O\7zCrc.obj \ + $O\7zCrcOpt.obj \ + $O\7zFile.obj \ + $O\7zDec.obj \ + $O\7zArcIn.obj \ + $O\7zStream.obj \ + $O\Bcj2.obj \ + $O\Bra.obj \ + $O\Bra86.obj \ + $O\BraIA64.obj \ + $O\CpuArch.obj \ + $O\Delta.obj \ + $O\Lzma2Dec.obj \ + $O\LzmaDec.obj \ + $O\Ppmd7.obj \ + $O\Ppmd7Dec.obj \ + +7Z_OBJS = \ + $O\7zMain.obj \ + +OBJS = \ + $O\Precomp.obj \ + $(7Z_OBJS) \ + $(C_OBJS) \ + +!include "../../../CPP/Build.mak" + +$(7Z_OBJS): $(*B).c + $(CCOMPL_USE) +$(C_OBJS): ../../$(*B).c + $(CCOMPL_USE) +$O\Precomp.obj: Precomp.c + $(CCOMPL_PCH) diff --git a/deps/LZMA-SDK/C/Util/7z/makefile.gcc b/deps/LZMA-SDK/C/Util/7z/makefile.gcc index d6ef9b2a7..4263d6755 100644 --- a/deps/LZMA-SDK/C/Util/7z/makefile.gcc +++ b/deps/LZMA-SDK/C/Util/7z/makefile.gcc @@ -1,34 +1,34 @@ -PROG = 7zdec - -LOCAL_FLAGS = -D_7ZIP_PPMD_SUPPPORT - -include ../../../CPP/7zip/LzmaDec_gcc.mak - - -OBJS = \ - $(LZMA_DEC_OPT_OBJS) \ - $O/Bcj2.o \ - $O/Bra.o \ - $O/Bra86.o \ - $O/BraIA64.o \ - $O/CpuArch.o \ - $O/Delta.o \ - $O/Lzma2Dec.o \ - $O/LzmaDec.o \ - $O/Ppmd7.o \ - $O/Ppmd7Dec.o \ - $O/7zCrc.o \ - $O/7zCrcOpt.o \ - $O/Sha256.o \ - $O/Sha256Opt.o \ - $O/7zAlloc.o \ - $O/7zArcIn.o \ - $O/7zBuf.o \ - $O/7zBuf2.o \ - $O/7zDec.o \ - $O/7zMain.o \ - $O/7zFile.o \ - $O/7zStream.o \ - - -include ../../7zip_gcc_c.mak +PROG = 7zdec + +LOCAL_FLAGS = -D_7ZIP_PPMD_SUPPPORT + +include ../../../CPP/7zip/LzmaDec_gcc.mak + + +OBJS = \ + $(LZMA_DEC_OPT_OBJS) \ + $O/Bcj2.o \ + $O/Bra.o \ + $O/Bra86.o \ + $O/BraIA64.o \ + $O/CpuArch.o \ + $O/Delta.o \ + $O/Lzma2Dec.o \ + $O/LzmaDec.o \ + $O/Ppmd7.o \ + $O/Ppmd7Dec.o \ + $O/7zCrc.o \ + $O/7zCrcOpt.o \ + $O/Sha256.o \ + $O/Sha256Opt.o \ + $O/7zAlloc.o \ + $O/7zArcIn.o \ + $O/7zBuf.o \ + $O/7zBuf2.o \ + $O/7zDec.o \ + $O/7zMain.o \ + $O/7zFile.o \ + $O/7zStream.o \ + + +include ../../7zip_gcc_c.mak diff --git a/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.c b/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.c index 4f05f1e73..62a59079f 100644 --- a/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.c +++ b/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.c @@ -1,286 +1,286 @@ -/* LzmaUtil.c -- Test application for LZMA compression -2021-11-01 : Igor Pavlov : Public domain */ - -#include "../../Precomp.h" - -#include -#include -#include - -#include "../../CpuArch.h" - -#include "../../Alloc.h" -#include "../../7zFile.h" -#include "../../7zVersion.h" -#include "../../LzFind.h" -#include "../../LzmaDec.h" -#include "../../LzmaEnc.h" - -static const char * const kCantReadMessage = "Cannot read input file"; -static const char * const kCantWriteMessage = "Cannot write output file"; -static const char * const kCantAllocateMessage = "Cannot allocate memory"; -static const char * const kDataErrorMessage = "Data error"; - -static void PrintHelp(char *buffer) -{ - strcat(buffer, - "\nLZMA-C " MY_VERSION_CPU " : " MY_COPYRIGHT_DATE "\n\n" - "Usage: lzma inputFile outputFile\n" - " e: encode file\n" - " d: decode file\n"); -} - -static int PrintError(char *buffer, const char *message) -{ - strcat(buffer, "\nError: "); - strcat(buffer, message); - strcat(buffer, "\n"); - return 1; -} - -static int PrintError_WRes(char *buffer, const char *message, WRes wres) -{ - strcat(buffer, "\nError: "); - strcat(buffer, message); - sprintf(buffer + strlen(buffer), "\nSystem error code: %d", (unsigned)wres); - #ifndef _WIN32 - { - const char *s = strerror(wres); - if (s) - sprintf(buffer + strlen(buffer), " : %s", s); - } - #endif - strcat(buffer, "\n"); - return 1; -} - -static int PrintErrorNumber(char *buffer, SRes val) -{ - sprintf(buffer + strlen(buffer), "\n7-Zip error code: %d\n", (unsigned)val); - return 1; -} - -static int PrintUserError(char *buffer) -{ - return PrintError(buffer, "Incorrect command"); -} - - -#define IN_BUF_SIZE (1 << 16) -#define OUT_BUF_SIZE (1 << 16) - - -static SRes Decode2(CLzmaDec *state, ISeqOutStream *outStream, ISeqInStream *inStream, - UInt64 unpackSize) -{ - int thereIsSize = (unpackSize != (UInt64)(Int64)-1); - Byte inBuf[IN_BUF_SIZE]; - Byte outBuf[OUT_BUF_SIZE]; - size_t inPos = 0, inSize = 0, outPos = 0; - LzmaDec_Init(state); - for (;;) - { - if (inPos == inSize) - { - inSize = IN_BUF_SIZE; - RINOK(inStream->Read(inStream, inBuf, &inSize)); - inPos = 0; - } - { - SRes res; - SizeT inProcessed = inSize - inPos; - SizeT outProcessed = OUT_BUF_SIZE - outPos; - ELzmaFinishMode finishMode = LZMA_FINISH_ANY; - ELzmaStatus status; - if (thereIsSize && outProcessed > unpackSize) - { - outProcessed = (SizeT)unpackSize; - finishMode = LZMA_FINISH_END; - } - - res = LzmaDec_DecodeToBuf(state, outBuf + outPos, &outProcessed, - inBuf + inPos, &inProcessed, finishMode, &status); - inPos += inProcessed; - outPos += outProcessed; - unpackSize -= outProcessed; - - if (outStream) - if (outStream->Write(outStream, outBuf, outPos) != outPos) - return SZ_ERROR_WRITE; - - outPos = 0; - - if (res != SZ_OK || (thereIsSize && unpackSize == 0)) - return res; - - if (inProcessed == 0 && outProcessed == 0) - { - if (thereIsSize || status != LZMA_STATUS_FINISHED_WITH_MARK) - return SZ_ERROR_DATA; - return res; - } - } - } -} - - -static SRes Decode(ISeqOutStream *outStream, ISeqInStream *inStream) -{ - UInt64 unpackSize; - int i; - SRes res = 0; - - CLzmaDec state; - - /* header: 5 bytes of LZMA properties and 8 bytes of uncompressed size */ - unsigned char header[LZMA_PROPS_SIZE + 8]; - - /* Read and parse header */ - - RINOK(SeqInStream_Read(inStream, header, sizeof(header))); - - unpackSize = 0; - for (i = 0; i < 8; i++) - unpackSize += (UInt64)header[LZMA_PROPS_SIZE + i] << (i * 8); - - LzmaDec_Construct(&state); - RINOK(LzmaDec_Allocate(&state, header, LZMA_PROPS_SIZE, &g_Alloc)); - res = Decode2(&state, outStream, inStream, unpackSize); - LzmaDec_Free(&state, &g_Alloc); - return res; -} - -static SRes Encode(ISeqOutStream *outStream, ISeqInStream *inStream, UInt64 fileSize, char *rs) -{ - CLzmaEncHandle enc; - SRes res; - CLzmaEncProps props; - - UNUSED_VAR(rs); - - enc = LzmaEnc_Create(&g_Alloc); - if (enc == 0) - return SZ_ERROR_MEM; - - LzmaEncProps_Init(&props); - res = LzmaEnc_SetProps(enc, &props); - - if (res == SZ_OK) - { - Byte header[LZMA_PROPS_SIZE + 8]; - size_t headerSize = LZMA_PROPS_SIZE; - int i; - - res = LzmaEnc_WriteProperties(enc, header, &headerSize); - for (i = 0; i < 8; i++) - header[headerSize++] = (Byte)(fileSize >> (8 * i)); - if (outStream->Write(outStream, header, headerSize) != headerSize) - res = SZ_ERROR_WRITE; - else - { - if (res == SZ_OK) - res = LzmaEnc_Encode(enc, outStream, inStream, NULL, &g_Alloc, &g_Alloc); - } - } - LzmaEnc_Destroy(enc, &g_Alloc, &g_Alloc); - return res; -} - - -static int main2(int numArgs, const char *args[], char *rs) -{ - CFileSeqInStream inStream; - CFileOutStream outStream; - char c; - int res; - int encodeMode; - BoolInt useOutFile = False; - - LzFindPrepare(); - - FileSeqInStream_CreateVTable(&inStream); - File_Construct(&inStream.file); - inStream.wres = 0; - - FileOutStream_CreateVTable(&outStream); - File_Construct(&outStream.file); - outStream.wres = 0; - - if (numArgs == 1) - { - PrintHelp(rs); - return 0; - } - - if (numArgs < 3 || numArgs > 4 || strlen(args[1]) != 1) - return PrintUserError(rs); - - c = args[1][0]; - encodeMode = (c == 'e' || c == 'E'); - if (!encodeMode && c != 'd' && c != 'D') - return PrintUserError(rs); - - { - size_t t4 = sizeof(UInt32); - size_t t8 = sizeof(UInt64); - if (t4 != 4 || t8 != 8) - return PrintError(rs, "Incorrect UInt32 or UInt64"); - } - - { - WRes wres = InFile_Open(&inStream.file, args[2]); - if (wres != 0) - return PrintError_WRes(rs, "Cannot open input file", wres); - } - - if (numArgs > 3) - { - WRes wres; - useOutFile = True; - wres = OutFile_Open(&outStream.file, args[3]); - if (wres != 0) - return PrintError_WRes(rs, "Cannot open output file", wres); - } - else if (encodeMode) - PrintUserError(rs); - - if (encodeMode) - { - UInt64 fileSize; - WRes wres = File_GetLength(&inStream.file, &fileSize); - if (wres != 0) - return PrintError_WRes(rs, "Cannot get file length", wres); - res = Encode(&outStream.vt, &inStream.vt, fileSize, rs); - } - else - { - res = Decode(&outStream.vt, useOutFile ? &inStream.vt : NULL); - } - - if (useOutFile) - File_Close(&outStream.file); - File_Close(&inStream.file); - - if (res != SZ_OK) - { - if (res == SZ_ERROR_MEM) - return PrintError(rs, kCantAllocateMessage); - else if (res == SZ_ERROR_DATA) - return PrintError(rs, kDataErrorMessage); - else if (res == SZ_ERROR_WRITE) - return PrintError_WRes(rs, kCantWriteMessage, outStream.wres); - else if (res == SZ_ERROR_READ) - return PrintError_WRes(rs, kCantReadMessage, inStream.wres); - return PrintErrorNumber(rs, res); - } - return 0; -} - - -int MY_CDECL main(int numArgs, const char *args[]) -{ - char rs[1000] = { 0 }; - int res = main2(numArgs, args, rs); - fputs(rs, stdout); - return res; -} +/* LzmaUtil.c -- Test application for LZMA compression +2021-11-01 : Igor Pavlov : Public domain */ + +#include "../../Precomp.h" + +#include +#include +#include + +#include "../../CpuArch.h" + +#include "../../Alloc.h" +#include "../../7zFile.h" +#include "../../7zVersion.h" +#include "../../LzFind.h" +#include "../../LzmaDec.h" +#include "../../LzmaEnc.h" + +static const char * const kCantReadMessage = "Cannot read input file"; +static const char * const kCantWriteMessage = "Cannot write output file"; +static const char * const kCantAllocateMessage = "Cannot allocate memory"; +static const char * const kDataErrorMessage = "Data error"; + +static void PrintHelp(char *buffer) +{ + strcat(buffer, + "\nLZMA-C " MY_VERSION_CPU " : " MY_COPYRIGHT_DATE "\n\n" + "Usage: lzma inputFile outputFile\n" + " e: encode file\n" + " d: decode file\n"); +} + +static int PrintError(char *buffer, const char *message) +{ + strcat(buffer, "\nError: "); + strcat(buffer, message); + strcat(buffer, "\n"); + return 1; +} + +static int PrintError_WRes(char *buffer, const char *message, WRes wres) +{ + strcat(buffer, "\nError: "); + strcat(buffer, message); + sprintf(buffer + strlen(buffer), "\nSystem error code: %d", (unsigned)wres); + #ifndef _WIN32 + { + const char *s = strerror(wres); + if (s) + sprintf(buffer + strlen(buffer), " : %s", s); + } + #endif + strcat(buffer, "\n"); + return 1; +} + +static int PrintErrorNumber(char *buffer, SRes val) +{ + sprintf(buffer + strlen(buffer), "\n7-Zip error code: %d\n", (unsigned)val); + return 1; +} + +static int PrintUserError(char *buffer) +{ + return PrintError(buffer, "Incorrect command"); +} + + +#define IN_BUF_SIZE (1 << 16) +#define OUT_BUF_SIZE (1 << 16) + + +static SRes Decode2(CLzmaDec *state, ISeqOutStream *outStream, ISeqInStream *inStream, + UInt64 unpackSize) +{ + int thereIsSize = (unpackSize != (UInt64)(Int64)-1); + Byte inBuf[IN_BUF_SIZE]; + Byte outBuf[OUT_BUF_SIZE]; + size_t inPos = 0, inSize = 0, outPos = 0; + LzmaDec_Init(state); + for (;;) + { + if (inPos == inSize) + { + inSize = IN_BUF_SIZE; + RINOK(inStream->Read(inStream, inBuf, &inSize)); + inPos = 0; + } + { + SRes res; + SizeT inProcessed = inSize - inPos; + SizeT outProcessed = OUT_BUF_SIZE - outPos; + ELzmaFinishMode finishMode = LZMA_FINISH_ANY; + ELzmaStatus status; + if (thereIsSize && outProcessed > unpackSize) + { + outProcessed = (SizeT)unpackSize; + finishMode = LZMA_FINISH_END; + } + + res = LzmaDec_DecodeToBuf(state, outBuf + outPos, &outProcessed, + inBuf + inPos, &inProcessed, finishMode, &status); + inPos += inProcessed; + outPos += outProcessed; + unpackSize -= outProcessed; + + if (outStream) + if (outStream->Write(outStream, outBuf, outPos) != outPos) + return SZ_ERROR_WRITE; + + outPos = 0; + + if (res != SZ_OK || (thereIsSize && unpackSize == 0)) + return res; + + if (inProcessed == 0 && outProcessed == 0) + { + if (thereIsSize || status != LZMA_STATUS_FINISHED_WITH_MARK) + return SZ_ERROR_DATA; + return res; + } + } + } +} + + +static SRes Decode(ISeqOutStream *outStream, ISeqInStream *inStream) +{ + UInt64 unpackSize; + int i; + SRes res = 0; + + CLzmaDec state; + + /* header: 5 bytes of LZMA properties and 8 bytes of uncompressed size */ + unsigned char header[LZMA_PROPS_SIZE + 8]; + + /* Read and parse header */ + + RINOK(SeqInStream_Read(inStream, header, sizeof(header))); + + unpackSize = 0; + for (i = 0; i < 8; i++) + unpackSize += (UInt64)header[LZMA_PROPS_SIZE + i] << (i * 8); + + LzmaDec_Construct(&state); + RINOK(LzmaDec_Allocate(&state, header, LZMA_PROPS_SIZE, &g_Alloc)); + res = Decode2(&state, outStream, inStream, unpackSize); + LzmaDec_Free(&state, &g_Alloc); + return res; +} + +static SRes Encode(ISeqOutStream *outStream, ISeqInStream *inStream, UInt64 fileSize, char *rs) +{ + CLzmaEncHandle enc; + SRes res; + CLzmaEncProps props; + + UNUSED_VAR(rs); + + enc = LzmaEnc_Create(&g_Alloc); + if (enc == 0) + return SZ_ERROR_MEM; + + LzmaEncProps_Init(&props); + res = LzmaEnc_SetProps(enc, &props); + + if (res == SZ_OK) + { + Byte header[LZMA_PROPS_SIZE + 8]; + size_t headerSize = LZMA_PROPS_SIZE; + int i; + + res = LzmaEnc_WriteProperties(enc, header, &headerSize); + for (i = 0; i < 8; i++) + header[headerSize++] = (Byte)(fileSize >> (8 * i)); + if (outStream->Write(outStream, header, headerSize) != headerSize) + res = SZ_ERROR_WRITE; + else + { + if (res == SZ_OK) + res = LzmaEnc_Encode(enc, outStream, inStream, NULL, &g_Alloc, &g_Alloc); + } + } + LzmaEnc_Destroy(enc, &g_Alloc, &g_Alloc); + return res; +} + + +static int main2(int numArgs, const char *args[], char *rs) +{ + CFileSeqInStream inStream; + CFileOutStream outStream; + char c; + int res; + int encodeMode; + BoolInt useOutFile = False; + + LzFindPrepare(); + + FileSeqInStream_CreateVTable(&inStream); + File_Construct(&inStream.file); + inStream.wres = 0; + + FileOutStream_CreateVTable(&outStream); + File_Construct(&outStream.file); + outStream.wres = 0; + + if (numArgs == 1) + { + PrintHelp(rs); + return 0; + } + + if (numArgs < 3 || numArgs > 4 || strlen(args[1]) != 1) + return PrintUserError(rs); + + c = args[1][0]; + encodeMode = (c == 'e' || c == 'E'); + if (!encodeMode && c != 'd' && c != 'D') + return PrintUserError(rs); + + { + size_t t4 = sizeof(UInt32); + size_t t8 = sizeof(UInt64); + if (t4 != 4 || t8 != 8) + return PrintError(rs, "Incorrect UInt32 or UInt64"); + } + + { + WRes wres = InFile_Open(&inStream.file, args[2]); + if (wres != 0) + return PrintError_WRes(rs, "Cannot open input file", wres); + } + + if (numArgs > 3) + { + WRes wres; + useOutFile = True; + wres = OutFile_Open(&outStream.file, args[3]); + if (wres != 0) + return PrintError_WRes(rs, "Cannot open output file", wres); + } + else if (encodeMode) + PrintUserError(rs); + + if (encodeMode) + { + UInt64 fileSize; + WRes wres = File_GetLength(&inStream.file, &fileSize); + if (wres != 0) + return PrintError_WRes(rs, "Cannot get file length", wres); + res = Encode(&outStream.vt, &inStream.vt, fileSize, rs); + } + else + { + res = Decode(&outStream.vt, useOutFile ? &inStream.vt : NULL); + } + + if (useOutFile) + File_Close(&outStream.file); + File_Close(&inStream.file); + + if (res != SZ_OK) + { + if (res == SZ_ERROR_MEM) + return PrintError(rs, kCantAllocateMessage); + else if (res == SZ_ERROR_DATA) + return PrintError(rs, kDataErrorMessage); + else if (res == SZ_ERROR_WRITE) + return PrintError_WRes(rs, kCantWriteMessage, outStream.wres); + else if (res == SZ_ERROR_READ) + return PrintError_WRes(rs, kCantReadMessage, inStream.wres); + return PrintErrorNumber(rs, res); + } + return 0; +} + + +int MY_CDECL main(int numArgs, const char *args[]) +{ + char rs[1000] = { 0 }; + int res = main2(numArgs, args, rs); + fputs(rs, stdout); + return res; +} diff --git a/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsp b/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsp index 44835d598..4e38e4a65 100644 --- a/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsp +++ b/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsp @@ -1,172 +1,172 @@ -# Microsoft Developer Studio Project File - Name="LzmaUtil" - Package Owner=<4> -# Microsoft Developer Studio Generated Build File, Format Version 6.00 -# ** DO NOT EDIT ** - -# TARGTYPE "Win32 (x86) Console Application" 0x0103 - -CFG=LzmaUtil - Win32 Debug -!MESSAGE This is not a valid makefile. To build this project using NMAKE, -!MESSAGE use the Export Makefile command and run -!MESSAGE -!MESSAGE NMAKE /f "LzmaUtil.mak". -!MESSAGE -!MESSAGE You can specify a configuration when running NMAKE -!MESSAGE by defining the macro CFG on the command line. For example: -!MESSAGE -!MESSAGE NMAKE /f "LzmaUtil.mak" CFG="LzmaUtil - Win32 Debug" -!MESSAGE -!MESSAGE Possible choices for configuration are: -!MESSAGE -!MESSAGE "LzmaUtil - Win32 Release" (based on "Win32 (x86) Console Application") -!MESSAGE "LzmaUtil - Win32 Debug" (based on "Win32 (x86) Console Application") -!MESSAGE - -# Begin Project -# PROP AllowPerConfigDependencies 0 -# PROP Scc_ProjName "" -# PROP Scc_LocalPath "" -CPP=cl.exe -RSC=rc.exe - -!IF "$(CFG)" == "LzmaUtil - Win32 Release" - -# PROP BASE Use_MFC 0 -# PROP BASE Use_Debug_Libraries 0 -# PROP BASE Output_Dir "Release" -# PROP BASE Intermediate_Dir "Release" -# PROP BASE Target_Dir "" -# PROP Use_MFC 0 -# PROP Use_Debug_Libraries 0 -# PROP Output_Dir "Release" -# PROP Intermediate_Dir "Release" -# PROP Ignore_Export_Lib 0 -# PROP Target_Dir "" -# ADD BASE CPP /nologo /W3 /GX /O2 /D "WIN32" /D "NDEBUG" /D "_CONSOLE" /D "_MBCS" /YX /FD /c -# ADD CPP /nologo /MT /W4 /WX /O2 /D "WIN32" /D "NDEBUG" /D "_CONSOLE" /D "_MBCS" /FD /c -# SUBTRACT CPP /YX -# ADD BASE RSC /l 0x419 /d "NDEBUG" -# ADD RSC /l 0x419 /d "NDEBUG" -BSC32=bscmake.exe -# ADD BASE BSC32 /nologo -# ADD BSC32 /nologo -LINK32=link.exe -# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /machine:I386 -# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /machine:I386 /out:"c:\util\7lzma.exe" - -!ELSEIF "$(CFG)" == "LzmaUtil - Win32 Debug" - -# PROP BASE Use_MFC 0 -# PROP BASE Use_Debug_Libraries 1 -# PROP BASE Output_Dir "Debug" -# PROP BASE Intermediate_Dir "Debug" -# PROP BASE Target_Dir "" -# PROP Use_MFC 0 -# PROP Use_Debug_Libraries 1 -# PROP Output_Dir "Debug" -# PROP Intermediate_Dir "Debug" -# PROP Ignore_Export_Lib 0 -# PROP Target_Dir "" -# ADD BASE CPP /nologo /W3 /Gm /GX /ZI /Od /D "WIN32" /D "_DEBUG" /D "_CONSOLE" /D "_MBCS" /YX /FD /GZ /c -# ADD CPP /nologo /MTd /W4 /WX /Gm /ZI /Od /D "WIN32" /D "_DEBUG" /D "_CONSOLE" /D "_MBCS" /FD /GZ /c -# SUBTRACT CPP /YX -# ADD BASE RSC /l 0x419 /d "_DEBUG" -# ADD RSC /l 0x419 /d "_DEBUG" -BSC32=bscmake.exe -# ADD BASE BSC32 /nologo -# ADD BSC32 /nologo -LINK32=link.exe -# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /debug /machine:I386 /pdbtype:sept -# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /debug /machine:I386 /out:"c:\util\7lzma.exe" /pdbtype:sept - -!ENDIF - -# Begin Target - -# Name "LzmaUtil - Win32 Release" -# Name "LzmaUtil - Win32 Debug" -# Begin Source File - -SOURCE=..\..\7zFile.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zFile.h -# End Source File -# Begin Source File - -SOURCE=..\..\7zStream.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zTypes.h -# End Source File -# Begin Source File - -SOURCE=..\..\7zVersion.h -# End Source File -# Begin Source File - -SOURCE=..\..\Alloc.c -# End Source File -# Begin Source File - -SOURCE=..\..\Alloc.h -# End Source File -# Begin Source File - -SOURCE=..\..\CpuArch.h -# End Source File -# Begin Source File - -SOURCE=..\..\LzFind.c -# End Source File -# Begin Source File - -SOURCE=..\..\LzFind.h -# End Source File -# Begin Source File - -SOURCE=..\..\LzFindMt.c -# End Source File -# Begin Source File - -SOURCE=..\..\LzFindMt.h -# End Source File -# Begin Source File - -SOURCE=..\..\LzFindOpt.c -# End Source File -# Begin Source File - -SOURCE=..\..\LzHash.h -# End Source File -# Begin Source File - -SOURCE=..\..\LzmaDec.c -# End Source File -# Begin Source File - -SOURCE=..\..\LzmaDec.h -# End Source File -# Begin Source File - -SOURCE=..\..\LzmaEnc.c -# End Source File -# Begin Source File - -SOURCE=..\..\LzmaEnc.h -# End Source File -# Begin Source File - -SOURCE=.\LzmaUtil.c -# End Source File -# Begin Source File - -SOURCE=..\..\Threads.c -# End Source File -# Begin Source File - -SOURCE=..\..\Threads.h -# End Source File -# End Target -# End Project +# Microsoft Developer Studio Project File - Name="LzmaUtil" - Package Owner=<4> +# Microsoft Developer Studio Generated Build File, Format Version 6.00 +# ** DO NOT EDIT ** + +# TARGTYPE "Win32 (x86) Console Application" 0x0103 + +CFG=LzmaUtil - Win32 Debug +!MESSAGE This is not a valid makefile. To build this project using NMAKE, +!MESSAGE use the Export Makefile command and run +!MESSAGE +!MESSAGE NMAKE /f "LzmaUtil.mak". +!MESSAGE +!MESSAGE You can specify a configuration when running NMAKE +!MESSAGE by defining the macro CFG on the command line. For example: +!MESSAGE +!MESSAGE NMAKE /f "LzmaUtil.mak" CFG="LzmaUtil - Win32 Debug" +!MESSAGE +!MESSAGE Possible choices for configuration are: +!MESSAGE +!MESSAGE "LzmaUtil - Win32 Release" (based on "Win32 (x86) Console Application") +!MESSAGE "LzmaUtil - Win32 Debug" (based on "Win32 (x86) Console Application") +!MESSAGE + +# Begin Project +# PROP AllowPerConfigDependencies 0 +# PROP Scc_ProjName "" +# PROP Scc_LocalPath "" +CPP=cl.exe +RSC=rc.exe + +!IF "$(CFG)" == "LzmaUtil - Win32 Release" + +# PROP BASE Use_MFC 0 +# PROP BASE Use_Debug_Libraries 0 +# PROP BASE Output_Dir "Release" +# PROP BASE Intermediate_Dir "Release" +# PROP BASE Target_Dir "" +# PROP Use_MFC 0 +# PROP Use_Debug_Libraries 0 +# PROP Output_Dir "Release" +# PROP Intermediate_Dir "Release" +# PROP Ignore_Export_Lib 0 +# PROP Target_Dir "" +# ADD BASE CPP /nologo /W3 /GX /O2 /D "WIN32" /D "NDEBUG" /D "_CONSOLE" /D "_MBCS" /YX /FD /c +# ADD CPP /nologo /MT /W4 /WX /O2 /D "WIN32" /D "NDEBUG" /D "_CONSOLE" /D "_MBCS" /FD /c +# SUBTRACT CPP /YX +# ADD BASE RSC /l 0x419 /d "NDEBUG" +# ADD RSC /l 0x419 /d "NDEBUG" +BSC32=bscmake.exe +# ADD BASE BSC32 /nologo +# ADD BSC32 /nologo +LINK32=link.exe +# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /machine:I386 +# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /machine:I386 /out:"c:\util\7lzma.exe" + +!ELSEIF "$(CFG)" == "LzmaUtil - Win32 Debug" + +# PROP BASE Use_MFC 0 +# PROP BASE Use_Debug_Libraries 1 +# PROP BASE Output_Dir "Debug" +# PROP BASE Intermediate_Dir "Debug" +# PROP BASE Target_Dir "" +# PROP Use_MFC 0 +# PROP Use_Debug_Libraries 1 +# PROP Output_Dir "Debug" +# PROP Intermediate_Dir "Debug" +# PROP Ignore_Export_Lib 0 +# PROP Target_Dir "" +# ADD BASE CPP /nologo /W3 /Gm /GX /ZI /Od /D "WIN32" /D "_DEBUG" /D "_CONSOLE" /D "_MBCS" /YX /FD /GZ /c +# ADD CPP /nologo /MTd /W4 /WX /Gm /ZI /Od /D "WIN32" /D "_DEBUG" /D "_CONSOLE" /D "_MBCS" /FD /GZ /c +# SUBTRACT CPP /YX +# ADD BASE RSC /l 0x419 /d "_DEBUG" +# ADD RSC /l 0x419 /d "_DEBUG" +BSC32=bscmake.exe +# ADD BASE BSC32 /nologo +# ADD BSC32 /nologo +LINK32=link.exe +# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /debug /machine:I386 /pdbtype:sept +# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:console /debug /machine:I386 /out:"c:\util\7lzma.exe" /pdbtype:sept + +!ENDIF + +# Begin Target + +# Name "LzmaUtil - Win32 Release" +# Name "LzmaUtil - Win32 Debug" +# Begin Source File + +SOURCE=..\..\7zFile.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zFile.h +# End Source File +# Begin Source File + +SOURCE=..\..\7zStream.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zTypes.h +# End Source File +# Begin Source File + +SOURCE=..\..\7zVersion.h +# End Source File +# Begin Source File + +SOURCE=..\..\Alloc.c +# End Source File +# Begin Source File + +SOURCE=..\..\Alloc.h +# End Source File +# Begin Source File + +SOURCE=..\..\CpuArch.h +# End Source File +# Begin Source File + +SOURCE=..\..\LzFind.c +# End Source File +# Begin Source File + +SOURCE=..\..\LzFind.h +# End Source File +# Begin Source File + +SOURCE=..\..\LzFindMt.c +# End Source File +# Begin Source File + +SOURCE=..\..\LzFindMt.h +# End Source File +# Begin Source File + +SOURCE=..\..\LzFindOpt.c +# End Source File +# Begin Source File + +SOURCE=..\..\LzHash.h +# End Source File +# Begin Source File + +SOURCE=..\..\LzmaDec.c +# End Source File +# Begin Source File + +SOURCE=..\..\LzmaDec.h +# End Source File +# Begin Source File + +SOURCE=..\..\LzmaEnc.c +# End Source File +# Begin Source File + +SOURCE=..\..\LzmaEnc.h +# End Source File +# Begin Source File + +SOURCE=.\LzmaUtil.c +# End Source File +# Begin Source File + +SOURCE=..\..\Threads.c +# End Source File +# Begin Source File + +SOURCE=..\..\Threads.h +# End Source File +# End Target +# End Project diff --git a/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsw b/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsw index f43548752..c52eaf6d0 100644 --- a/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsw +++ b/deps/LZMA-SDK/C/Util/Lzma/LzmaUtil.dsw @@ -1,29 +1,29 @@ -Microsoft Developer Studio Workspace File, Format Version 6.00 -# WARNING: DO NOT EDIT OR DELETE THIS WORKSPACE FILE! - -############################################################################### - -Project: "LzmaUtil"=.\LzmaUtil.dsp - Package Owner=<4> - -Package=<5> -{{{ -}}} - -Package=<4> -{{{ -}}} - -############################################################################### - -Global: - -Package=<5> -{{{ -}}} - -Package=<3> -{{{ -}}} - -############################################################################### - +Microsoft Developer Studio Workspace File, Format Version 6.00 +# WARNING: DO NOT EDIT OR DELETE THIS WORKSPACE FILE! + +############################################################################### + +Project: "LzmaUtil"=.\LzmaUtil.dsp - Package Owner=<4> + +Package=<5> +{{{ +}}} + +Package=<4> +{{{ +}}} + +############################################################################### + +Global: + +Package=<5> +{{{ +}}} + +Package=<3> +{{{ +}}} + +############################################################################### + diff --git a/deps/LZMA-SDK/C/Util/Lzma/makefile b/deps/LZMA-SDK/C/Util/Lzma/makefile index 9d971f666..7813bdb05 100644 --- a/deps/LZMA-SDK/C/Util/Lzma/makefile +++ b/deps/LZMA-SDK/C/Util/Lzma/makefile @@ -1,30 +1,30 @@ -# MY_STATIC_LINK=1 -PROG = LZMAc.exe - -CFLAGS = $(CFLAGS) \ - -LIB_OBJS = \ - $O\LzmaUtil.obj \ - -C_OBJS = \ - $O\Alloc.obj \ - $O\CpuArch.obj \ - $O\LzFind.obj \ - $O\LzFindMt.obj \ - $O\LzFindOpt.obj \ - $O\LzmaDec.obj \ - $O\LzmaEnc.obj \ - $O\7zFile.obj \ - $O\7zStream.obj \ - $O\Threads.obj \ - -OBJS = \ - $(LIB_OBJS) \ - $(C_OBJS) \ - -!include "../../../CPP/Build.mak" - -$(LIB_OBJS): $(*B).c - $(COMPL_O2) -$(C_OBJS): ../../$(*B).c - $(COMPL_O2) +# MY_STATIC_LINK=1 +PROG = LZMAc.exe + +CFLAGS = $(CFLAGS) \ + +LIB_OBJS = \ + $O\LzmaUtil.obj \ + +C_OBJS = \ + $O\Alloc.obj \ + $O\CpuArch.obj \ + $O\LzFind.obj \ + $O\LzFindMt.obj \ + $O\LzFindOpt.obj \ + $O\LzmaDec.obj \ + $O\LzmaEnc.obj \ + $O\7zFile.obj \ + $O\7zStream.obj \ + $O\Threads.obj \ + +OBJS = \ + $(LIB_OBJS) \ + $(C_OBJS) \ + +!include "../../../CPP/Build.mak" + +$(LIB_OBJS): $(*B).c + $(COMPL_O2) +$(C_OBJS): ../../$(*B).c + $(COMPL_O2) diff --git a/deps/LZMA-SDK/C/Util/Lzma/makefile.gcc b/deps/LZMA-SDK/C/Util/Lzma/makefile.gcc index 6ce77a0f2..2acb0b808 100644 --- a/deps/LZMA-SDK/C/Util/Lzma/makefile.gcc +++ b/deps/LZMA-SDK/C/Util/Lzma/makefile.gcc @@ -1,21 +1,21 @@ -PROG = 7lzma - -include ../../../CPP/7zip/LzmaDec_gcc.mak - - -OBJS = \ - $(LZMA_DEC_OPT_OBJS) \ - $O/7zFile.o \ - $O/7zStream.o \ - $O/Alloc.o \ - $O/CpuArch.o \ - $O/LzFind.o \ - $O/LzFindMt.o \ - $O/LzFindOpt.o \ - $O/LzmaDec.o \ - $O/LzmaEnc.o \ - $O/LzmaUtil.o \ - $O/Threads.o \ - - -include ../../7zip_gcc_c.mak +PROG = 7lzma + +include ../../../CPP/7zip/LzmaDec_gcc.mak + + +OBJS = \ + $(LZMA_DEC_OPT_OBJS) \ + $O/7zFile.o \ + $O/7zStream.o \ + $O/Alloc.o \ + $O/CpuArch.o \ + $O/LzFind.o \ + $O/LzFindMt.o \ + $O/LzFindOpt.o \ + $O/LzmaDec.o \ + $O/LzmaEnc.o \ + $O/LzmaUtil.o \ + $O/Threads.o \ + + +include ../../7zip_gcc_c.mak diff --git a/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.def b/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.def index 43b959778..8bc6add93 100644 --- a/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.def +++ b/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.def @@ -1,4 +1,4 @@ -EXPORTS - LzmaCompress - LzmaUncompress - +EXPORTS + LzmaCompress + LzmaUncompress + diff --git a/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsp b/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsp index c267bbef6..6ce91dcd2 100644 --- a/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsp +++ b/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsp @@ -1,182 +1,182 @@ -# Microsoft Developer Studio Project File - Name="LzmaLib" - Package Owner=<4> -# Microsoft Developer Studio Generated Build File, Format Version 6.00 -# ** DO NOT EDIT ** - -# TARGTYPE "Win32 (x86) Dynamic-Link Library" 0x0102 - -CFG=LzmaLib - Win32 Debug -!MESSAGE This is not a valid makefile. To build this project using NMAKE, -!MESSAGE use the Export Makefile command and run -!MESSAGE -!MESSAGE NMAKE /f "LzmaLib.mak". -!MESSAGE -!MESSAGE You can specify a configuration when running NMAKE -!MESSAGE by defining the macro CFG on the command line. For example: -!MESSAGE -!MESSAGE NMAKE /f "LzmaLib.mak" CFG="LzmaLib - Win32 Debug" -!MESSAGE -!MESSAGE Possible choices for configuration are: -!MESSAGE -!MESSAGE "LzmaLib - Win32 Release" (based on "Win32 (x86) Dynamic-Link Library") -!MESSAGE "LzmaLib - Win32 Debug" (based on "Win32 (x86) Dynamic-Link Library") -!MESSAGE - -# Begin Project -# PROP AllowPerConfigDependencies 0 -# PROP Scc_ProjName "" -# PROP Scc_LocalPath "" -CPP=cl.exe -MTL=midl.exe -RSC=rc.exe - -!IF "$(CFG)" == "LzmaLib - Win32 Release" - -# PROP BASE Use_MFC 0 -# PROP BASE Use_Debug_Libraries 0 -# PROP BASE Output_Dir "Release" -# PROP BASE Intermediate_Dir "Release" -# PROP BASE Target_Dir "" -# PROP Use_MFC 0 -# PROP Use_Debug_Libraries 0 -# PROP Output_Dir "Release" -# PROP Intermediate_Dir "Release" -# PROP Ignore_Export_Lib 0 -# PROP Target_Dir "" -# ADD BASE CPP /nologo /MT /W3 /GX /O2 /D "WIN32" /D "NDEBUG" /D "_WINDOWS" /D "_MBCS" /D "_USRDLL" /D "LZMALIB_EXPORTS" /YX /FD /c -# ADD CPP /nologo /Gr /MT /W3 /O2 /D "NDEBUG" /D "WIN32" /D "_WINDOWS" /D "_MBCS" /D "_USRDLL" /D "LZMALIB_EXPORTS" /FD /c -# SUBTRACT CPP /YX -# ADD BASE MTL /nologo /D "NDEBUG" /mktyplib203 /win32 -# ADD MTL /nologo /D "NDEBUG" /mktyplib203 /win32 -# ADD BASE RSC /l 0x419 /d "NDEBUG" -# ADD RSC /l 0x419 /d "NDEBUG" -BSC32=bscmake.exe -# ADD BASE BSC32 /nologo -# ADD BSC32 /nologo -LINK32=link.exe -# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /dll /machine:I386 -# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /dll /machine:I386 /out:"C:\Util\LZMA.dll" /opt:NOWIN98 -# SUBTRACT LINK32 /pdb:none - -!ELSEIF "$(CFG)" == "LzmaLib - Win32 Debug" - -# PROP BASE Use_MFC 0 -# PROP BASE Use_Debug_Libraries 1 -# PROP BASE Output_Dir "Debug" -# PROP BASE Intermediate_Dir "Debug" -# PROP BASE Target_Dir "" -# PROP Use_MFC 0 -# PROP Use_Debug_Libraries 1 -# PROP Output_Dir "Debug" -# PROP Intermediate_Dir "Debug" -# PROP Ignore_Export_Lib 0 -# PROP Target_Dir "" -# ADD BASE CPP /nologo /MTd /W3 /Gm /GX /ZI /Od /D "WIN32" /D "_DEBUG" /D "_WINDOWS" /D "_MBCS" /D "_USRDLL" /D "LZMALIB_EXPORTS" /YX /FD /GZ /c -# ADD CPP /nologo /MTd /W3 /Gm /ZI /Od /D "_DEBUG" /D "WIN32" /D "_WINDOWS" /D "_MBCS" /D "_USRDLL" /D "LZMALIB_EXPORTS" /D "COMPRESS_MF_MT" /FD /GZ /c -# SUBTRACT CPP /YX -# ADD BASE MTL /nologo /D "_DEBUG" /mktyplib203 /win32 -# ADD MTL /nologo /D "_DEBUG" /mktyplib203 /win32 -# ADD BASE RSC /l 0x419 /d "_DEBUG" -# ADD RSC /l 0x419 /d "_DEBUG" -BSC32=bscmake.exe -# ADD BASE BSC32 /nologo -# ADD BSC32 /nologo -LINK32=link.exe -# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /dll /debug /machine:I386 /pdbtype:sept -# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /dll /debug /machine:I386 /out:"C:\Util\LZMA.dll" /pdbtype:sept - -!ENDIF - -# Begin Target - -# Name "LzmaLib - Win32 Release" -# Name "LzmaLib - Win32 Debug" -# Begin Group "Spec" - -# PROP Default_Filter "" -# Begin Source File - -SOURCE=.\LzmaLib.def -# End Source File -# Begin Source File - -SOURCE=.\LzmaLibExports.c -# End Source File -# End Group -# Begin Source File - -SOURCE=..\..\7zTypes.h -# End Source File -# Begin Source File - -SOURCE=..\..\Alloc.c -# End Source File -# Begin Source File - -SOURCE=..\..\Alloc.h -# End Source File -# Begin Source File - -SOURCE=..\..\IStream.h -# End Source File -# Begin Source File - -SOURCE=..\..\LzFind.c -# End Source File -# Begin Source File - -SOURCE=..\..\LzFind.h -# End Source File -# Begin Source File - -SOURCE=..\..\LzFindMt.c -# End Source File -# Begin Source File - -SOURCE=..\..\LzFindMt.h -# End Source File -# Begin Source File - -SOURCE=..\..\LzFindOpt.c -# End Source File -# Begin Source File - -SOURCE=..\..\LzHash.h -# End Source File -# Begin Source File - -SOURCE=..\..\LzmaDec.c -# End Source File -# Begin Source File - -SOURCE=..\..\LzmaDec.h -# End Source File -# Begin Source File - -SOURCE=..\..\LzmaEnc.c -# End Source File -# Begin Source File - -SOURCE=..\..\LzmaEnc.h -# End Source File -# Begin Source File - -SOURCE=..\..\LzmaLib.c -# End Source File -# Begin Source File - -SOURCE=..\..\LzmaLib.h -# End Source File -# Begin Source File - -SOURCE=.\resource.rc -# End Source File -# Begin Source File - -SOURCE=..\..\Threads.c -# End Source File -# Begin Source File - -SOURCE=..\..\Threads.h -# End Source File -# End Target -# End Project +# Microsoft Developer Studio Project File - Name="LzmaLib" - Package Owner=<4> +# Microsoft Developer Studio Generated Build File, Format Version 6.00 +# ** DO NOT EDIT ** + +# TARGTYPE "Win32 (x86) Dynamic-Link Library" 0x0102 + +CFG=LzmaLib - Win32 Debug +!MESSAGE This is not a valid makefile. To build this project using NMAKE, +!MESSAGE use the Export Makefile command and run +!MESSAGE +!MESSAGE NMAKE /f "LzmaLib.mak". +!MESSAGE +!MESSAGE You can specify a configuration when running NMAKE +!MESSAGE by defining the macro CFG on the command line. For example: +!MESSAGE +!MESSAGE NMAKE /f "LzmaLib.mak" CFG="LzmaLib - Win32 Debug" +!MESSAGE +!MESSAGE Possible choices for configuration are: +!MESSAGE +!MESSAGE "LzmaLib - Win32 Release" (based on "Win32 (x86) Dynamic-Link Library") +!MESSAGE "LzmaLib - Win32 Debug" (based on "Win32 (x86) Dynamic-Link Library") +!MESSAGE + +# Begin Project +# PROP AllowPerConfigDependencies 0 +# PROP Scc_ProjName "" +# PROP Scc_LocalPath "" +CPP=cl.exe +MTL=midl.exe +RSC=rc.exe + +!IF "$(CFG)" == "LzmaLib - Win32 Release" + +# PROP BASE Use_MFC 0 +# PROP BASE Use_Debug_Libraries 0 +# PROP BASE Output_Dir "Release" +# PROP BASE Intermediate_Dir "Release" +# PROP BASE Target_Dir "" +# PROP Use_MFC 0 +# PROP Use_Debug_Libraries 0 +# PROP Output_Dir "Release" +# PROP Intermediate_Dir "Release" +# PROP Ignore_Export_Lib 0 +# PROP Target_Dir "" +# ADD BASE CPP /nologo /MT /W3 /GX /O2 /D "WIN32" /D "NDEBUG" /D "_WINDOWS" /D "_MBCS" /D "_USRDLL" /D "LZMALIB_EXPORTS" /YX /FD /c +# ADD CPP /nologo /Gr /MT /W3 /O2 /D "NDEBUG" /D "WIN32" /D "_WINDOWS" /D "_MBCS" /D "_USRDLL" /D "LZMALIB_EXPORTS" /FD /c +# SUBTRACT CPP /YX +# ADD BASE MTL /nologo /D "NDEBUG" /mktyplib203 /win32 +# ADD MTL /nologo /D "NDEBUG" /mktyplib203 /win32 +# ADD BASE RSC /l 0x419 /d "NDEBUG" +# ADD RSC /l 0x419 /d "NDEBUG" +BSC32=bscmake.exe +# ADD BASE BSC32 /nologo +# ADD BSC32 /nologo +LINK32=link.exe +# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /dll /machine:I386 +# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /dll /machine:I386 /out:"C:\Util\LZMA.dll" /opt:NOWIN98 +# SUBTRACT LINK32 /pdb:none + +!ELSEIF "$(CFG)" == "LzmaLib - Win32 Debug" + +# PROP BASE Use_MFC 0 +# PROP BASE Use_Debug_Libraries 1 +# PROP BASE Output_Dir "Debug" +# PROP BASE Intermediate_Dir "Debug" +# PROP BASE Target_Dir "" +# PROP Use_MFC 0 +# PROP Use_Debug_Libraries 1 +# PROP Output_Dir "Debug" +# PROP Intermediate_Dir "Debug" +# PROP Ignore_Export_Lib 0 +# PROP Target_Dir "" +# ADD BASE CPP /nologo /MTd /W3 /Gm /GX /ZI /Od /D "WIN32" /D "_DEBUG" /D "_WINDOWS" /D "_MBCS" /D "_USRDLL" /D "LZMALIB_EXPORTS" /YX /FD /GZ /c +# ADD CPP /nologo /MTd /W3 /Gm /ZI /Od /D "_DEBUG" /D "WIN32" /D "_WINDOWS" /D "_MBCS" /D "_USRDLL" /D "LZMALIB_EXPORTS" /D "COMPRESS_MF_MT" /FD /GZ /c +# SUBTRACT CPP /YX +# ADD BASE MTL /nologo /D "_DEBUG" /mktyplib203 /win32 +# ADD MTL /nologo /D "_DEBUG" /mktyplib203 /win32 +# ADD BASE RSC /l 0x419 /d "_DEBUG" +# ADD RSC /l 0x419 /d "_DEBUG" +BSC32=bscmake.exe +# ADD BASE BSC32 /nologo +# ADD BSC32 /nologo +LINK32=link.exe +# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /dll /debug /machine:I386 /pdbtype:sept +# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /dll /debug /machine:I386 /out:"C:\Util\LZMA.dll" /pdbtype:sept + +!ENDIF + +# Begin Target + +# Name "LzmaLib - Win32 Release" +# Name "LzmaLib - Win32 Debug" +# Begin Group "Spec" + +# PROP Default_Filter "" +# Begin Source File + +SOURCE=.\LzmaLib.def +# End Source File +# Begin Source File + +SOURCE=.\LzmaLibExports.c +# End Source File +# End Group +# Begin Source File + +SOURCE=..\..\7zTypes.h +# End Source File +# Begin Source File + +SOURCE=..\..\Alloc.c +# End Source File +# Begin Source File + +SOURCE=..\..\Alloc.h +# End Source File +# Begin Source File + +SOURCE=..\..\IStream.h +# End Source File +# Begin Source File + +SOURCE=..\..\LzFind.c +# End Source File +# Begin Source File + +SOURCE=..\..\LzFind.h +# End Source File +# Begin Source File + +SOURCE=..\..\LzFindMt.c +# End Source File +# Begin Source File + +SOURCE=..\..\LzFindMt.h +# End Source File +# Begin Source File + +SOURCE=..\..\LzFindOpt.c +# End Source File +# Begin Source File + +SOURCE=..\..\LzHash.h +# End Source File +# Begin Source File + +SOURCE=..\..\LzmaDec.c +# End Source File +# Begin Source File + +SOURCE=..\..\LzmaDec.h +# End Source File +# Begin Source File + +SOURCE=..\..\LzmaEnc.c +# End Source File +# Begin Source File + +SOURCE=..\..\LzmaEnc.h +# End Source File +# Begin Source File + +SOURCE=..\..\LzmaLib.c +# End Source File +# Begin Source File + +SOURCE=..\..\LzmaLib.h +# End Source File +# Begin Source File + +SOURCE=.\resource.rc +# End Source File +# Begin Source File + +SOURCE=..\..\Threads.c +# End Source File +# Begin Source File + +SOURCE=..\..\Threads.h +# End Source File +# End Target +# End Project diff --git a/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsw b/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsw index f6c55593f..6faf33365 100644 --- a/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsw +++ b/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLib.dsw @@ -1,29 +1,29 @@ -Microsoft Developer Studio Workspace File, Format Version 6.00 -# WARNING: DO NOT EDIT OR DELETE THIS WORKSPACE FILE! - -############################################################################### - -Project: "LzmaLib"=.\LzmaLib.dsp - Package Owner=<4> - -Package=<5> -{{{ -}}} - -Package=<4> -{{{ -}}} - -############################################################################### - -Global: - -Package=<5> -{{{ -}}} - -Package=<3> -{{{ -}}} - -############################################################################### - +Microsoft Developer Studio Workspace File, Format Version 6.00 +# WARNING: DO NOT EDIT OR DELETE THIS WORKSPACE FILE! + +############################################################################### + +Project: "LzmaLib"=.\LzmaLib.dsp - Package Owner=<4> + +Package=<5> +{{{ +}}} + +Package=<4> +{{{ +}}} + +############################################################################### + +Global: + +Package=<5> +{{{ +}}} + +Package=<3> +{{{ +}}} + +############################################################################### + diff --git a/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLibExports.c b/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLibExports.c index 02600c724..4a28a9a6b 100644 --- a/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLibExports.c +++ b/deps/LZMA-SDK/C/Util/LzmaLib/LzmaLibExports.c @@ -1,14 +1,14 @@ -/* LzmaLibExports.c -- LZMA library DLL Entry point -2015-11-08 : Igor Pavlov : Public domain */ - -#include "../../Precomp.h" - -#include - -BOOL WINAPI DllMain(HINSTANCE hInstance, DWORD dwReason, LPVOID lpReserved) -{ - UNUSED_VAR(hInstance); - UNUSED_VAR(dwReason); - UNUSED_VAR(lpReserved); - return TRUE; -} +/* LzmaLibExports.c -- LZMA library DLL Entry point +2015-11-08 : Igor Pavlov : Public domain */ + +#include "../../Precomp.h" + +#include + +BOOL WINAPI DllMain(HINSTANCE hInstance, DWORD dwReason, LPVOID lpReserved) +{ + UNUSED_VAR(hInstance); + UNUSED_VAR(dwReason); + UNUSED_VAR(lpReserved); + return TRUE; +} diff --git a/deps/LZMA-SDK/C/Util/LzmaLib/makefile b/deps/LZMA-SDK/C/Util/LzmaLib/makefile index bcb7496be..b36f1de09 100644 --- a/deps/LZMA-SDK/C/Util/LzmaLib/makefile +++ b/deps/LZMA-SDK/C/Util/LzmaLib/makefile @@ -1,36 +1,36 @@ -MY_STATIC_LINK=1 -SLIB = sLZMA.lib -PROG = LZMA.dll -SLIBPATH = $O\$(SLIB) - -DEF_FILE = LzmaLib.def -CFLAGS = $(CFLAGS) \ - -LIB_OBJS = \ - $O\LzmaLibExports.obj \ - -C_OBJS = \ - $O\Alloc.obj \ - $O\CpuArch.obj \ - $O\LzFind.obj \ - $O\LzFindMt.obj \ - $O\LzFindOpt.obj \ - $O\LzmaDec.obj \ - $O\LzmaEnc.obj \ - $O\LzmaLib.obj \ - $O\Threads.obj \ - -OBJS = \ - $(LIB_OBJS) \ - $(C_OBJS) \ - $O\resource.res - -!include "../../../CPP/Build.mak" - -$(SLIBPATH): $O $(OBJS) - lib -out:$(SLIBPATH) $(OBJS) $(LIBS) - -$(LIB_OBJS): $(*B).c - $(COMPL_O2) -$(C_OBJS): ../../$(*B).c - $(COMPL_O2) +MY_STATIC_LINK=1 +SLIB = sLZMA.lib +PROG = LZMA.dll +SLIBPATH = $O\$(SLIB) + +DEF_FILE = LzmaLib.def +CFLAGS = $(CFLAGS) \ + +LIB_OBJS = \ + $O\LzmaLibExports.obj \ + +C_OBJS = \ + $O\Alloc.obj \ + $O\CpuArch.obj \ + $O\LzFind.obj \ + $O\LzFindMt.obj \ + $O\LzFindOpt.obj \ + $O\LzmaDec.obj \ + $O\LzmaEnc.obj \ + $O\LzmaLib.obj \ + $O\Threads.obj \ + +OBJS = \ + $(LIB_OBJS) \ + $(C_OBJS) \ + $O\resource.res + +!include "../../../CPP/Build.mak" + +$(SLIBPATH): $O $(OBJS) + lib -out:$(SLIBPATH) $(OBJS) $(LIBS) + +$(LIB_OBJS): $(*B).c + $(COMPL_O2) +$(C_OBJS): ../../$(*B).c + $(COMPL_O2) diff --git a/deps/LZMA-SDK/C/Util/LzmaLib/resource.rc b/deps/LZMA-SDK/C/Util/LzmaLib/resource.rc index d95e3f358..674832e05 100644 --- a/deps/LZMA-SDK/C/Util/LzmaLib/resource.rc +++ b/deps/LZMA-SDK/C/Util/LzmaLib/resource.rc @@ -1,3 +1,3 @@ -#include "../../7zVersion.rc" - -MY_VERSION_INFO_DLL("LZMA library", "LZMA") +#include "../../7zVersion.rc" + +MY_VERSION_INFO_DLL("LZMA library", "LZMA") diff --git a/deps/LZMA-SDK/C/Util/SfxSetup/Precomp.c b/deps/LZMA-SDK/C/Util/SfxSetup/Precomp.c index 34b60f8fc..01605e3c2 100644 --- a/deps/LZMA-SDK/C/Util/SfxSetup/Precomp.c +++ b/deps/LZMA-SDK/C/Util/SfxSetup/Precomp.c @@ -1,4 +1,4 @@ -/* Precomp.c -- StdAfx -2013-01-21 : Igor Pavlov : Public domain */ - -#include "Precomp.h" +/* Precomp.c -- StdAfx +2013-01-21 : Igor Pavlov : Public domain */ + +#include "Precomp.h" diff --git a/deps/LZMA-SDK/C/Util/SfxSetup/Precomp.h b/deps/LZMA-SDK/C/Util/SfxSetup/Precomp.h index 9f398d08f..588a66f7e 100644 --- a/deps/LZMA-SDK/C/Util/SfxSetup/Precomp.h +++ b/deps/LZMA-SDK/C/Util/SfxSetup/Precomp.h @@ -1,10 +1,10 @@ -/* Precomp.h -- StdAfx -2013-06-16 : Igor Pavlov : Public domain */ - -#ifndef __7Z_PRECOMP_H -#define __7Z_PRECOMP_H - -#include "../../Compiler.h" -#include "../../7zTypes.h" - -#endif +/* Precomp.h -- StdAfx +2013-06-16 : Igor Pavlov : Public domain */ + +#ifndef __7Z_PRECOMP_H +#define __7Z_PRECOMP_H + +#include "../../Compiler.h" +#include "../../7zTypes.h" + +#endif diff --git a/deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.c b/deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.c index 394369abd..ef19aeac5 100644 --- a/deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.c +++ b/deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.c @@ -1,640 +1,640 @@ -/* SfxSetup.c - 7z SFX Setup -2019-02-02 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#ifndef UNICODE -#define UNICODE -#endif - -#ifndef _UNICODE -#define _UNICODE -#endif - -#ifdef _CONSOLE -#include -#endif - -#include "../../7z.h" -#include "../../7zAlloc.h" -#include "../../7zCrc.h" -#include "../../7zFile.h" -#include "../../CpuArch.h" -#include "../../DllSecur.h" - -#define k_EXE_ExtIndex 2 - -#define kInputBufSize ((size_t)1 << 18) - -static const char * const kExts[] = -{ - "bat" - , "cmd" - , "exe" - , "inf" - , "msi" - #ifdef UNDER_CE - , "cab" - #endif - , "html" - , "htm" -}; - -static const char * const kNames[] = -{ - "setup" - , "install" - , "run" - , "start" -}; - -static unsigned FindExt(const wchar_t *s, unsigned *extLen) -{ - unsigned len = (unsigned)wcslen(s); - unsigned i; - for (i = len; i > 0; i--) - { - if (s[i - 1] == '.') - { - *extLen = len - i; - return i - 1; - } - } - *extLen = 0; - return len; -} - -#define MAKE_CHAR_UPPER(c) ((((c) >= 'a' && (c) <= 'z') ? (c) -= 0x20 : (c))) - -static unsigned FindItem(const char * const *items, unsigned num, const wchar_t *s, unsigned len) -{ - unsigned i; - for (i = 0; i < num; i++) - { - const char *item = items[i]; - unsigned itemLen = (unsigned)strlen(item); - unsigned j; - if (len != itemLen) - continue; - for (j = 0; j < len; j++) - { - unsigned c = (Byte)item[j]; - if (c != s[j] && MAKE_CHAR_UPPER(c) != s[j]) - break; - } - if (j == len) - return i; - } - return i; -} - -#ifdef _CONSOLE -static BOOL WINAPI HandlerRoutine(DWORD ctrlType) -{ - UNUSED_VAR(ctrlType); - return TRUE; -} -#endif - -static void PrintErrorMessage(const char *message) -{ - #ifdef _CONSOLE - printf("\n7-Zip Error: %s\n", message); - #else - #ifdef UNDER_CE - WCHAR messageW[256 + 4]; - unsigned i; - for (i = 0; i < 256 && message[i] != 0; i++) - messageW[i] = message[i]; - messageW[i] = 0; - MessageBoxW(0, messageW, L"7-Zip Error", MB_ICONERROR); - #else - MessageBoxA(0, message, "7-Zip Error", MB_ICONERROR); - #endif - #endif -} - -static WRes MyCreateDir(const WCHAR *name) -{ - return CreateDirectoryW(name, NULL) ? 0 : GetLastError(); -} - -#ifdef UNDER_CE -#define kBufferSize (1 << 13) -#else -#define kBufferSize (1 << 15) -#endif - -#define kSignatureSearchLimit (1 << 22) - -static BoolInt FindSignature(CSzFile *stream, UInt64 *resPos) -{ - Byte buf[kBufferSize]; - size_t numPrevBytes = 0; - *resPos = 0; - for (;;) - { - size_t processed, pos; - if (*resPos > kSignatureSearchLimit) - return False; - processed = kBufferSize - numPrevBytes; - if (File_Read(stream, buf + numPrevBytes, &processed) != 0) - return False; - processed += numPrevBytes; - if (processed < k7zStartHeaderSize || - (processed == k7zStartHeaderSize && numPrevBytes != 0)) - return False; - processed -= k7zStartHeaderSize; - for (pos = 0; pos <= processed; pos++) - { - for (; pos <= processed && buf[pos] != '7'; pos++); - if (pos > processed) - break; - if (memcmp(buf + pos, k7zSignature, k7zSignatureSize) == 0) - if (CrcCalc(buf + pos + 12, 20) == GetUi32(buf + pos + 8)) - { - *resPos += pos; - return True; - } - } - *resPos += processed; - numPrevBytes = k7zStartHeaderSize; - memmove(buf, buf + processed, k7zStartHeaderSize); - } -} - -static BoolInt DoesFileOrDirExist(const WCHAR *path) -{ - WIN32_FIND_DATAW fd; - HANDLE handle; - handle = FindFirstFileW(path, &fd); - if (handle == INVALID_HANDLE_VALUE) - return False; - FindClose(handle); - return True; -} - -static WRes RemoveDirWithSubItems(WCHAR *path) -{ - WIN32_FIND_DATAW fd; - HANDLE handle; - WRes res = 0; - size_t len = wcslen(path); - wcscpy(path + len, L"*"); - handle = FindFirstFileW(path, &fd); - path[len] = L'\0'; - if (handle == INVALID_HANDLE_VALUE) - return GetLastError(); - - for (;;) - { - if (wcscmp(fd.cFileName, L".") != 0 && - wcscmp(fd.cFileName, L"..") != 0) - { - wcscpy(path + len, fd.cFileName); - if ((fd.dwFileAttributes & FILE_ATTRIBUTE_DIRECTORY) != 0) - { - wcscat(path, WSTRING_PATH_SEPARATOR); - res = RemoveDirWithSubItems(path); - } - else - { - SetFileAttributesW(path, 0); - if (DeleteFileW(path) == 0) - res = GetLastError(); - } - - if (res != 0) - break; - } - - if (!FindNextFileW(handle, &fd)) - { - res = GetLastError(); - if (res == ERROR_NO_MORE_FILES) - res = 0; - break; - } - } - - path[len] = L'\0'; - FindClose(handle); - if (res == 0) - { - if (!RemoveDirectoryW(path)) - res = GetLastError(); - } - return res; -} - -#ifdef _CONSOLE -int MY_CDECL main() -#else -int APIENTRY WinMain(HINSTANCE hInstance, HINSTANCE hPrevInstance, - #ifdef UNDER_CE - LPWSTR - #else - LPSTR - #endif - lpCmdLine, int nCmdShow) -#endif -{ - CFileInStream archiveStream; - CLookToRead2 lookStream; - CSzArEx db; - SRes res = SZ_OK; - ISzAlloc allocImp; - ISzAlloc allocTempImp; - WCHAR sfxPath[MAX_PATH + 2]; - WCHAR path[MAX_PATH * 3 + 2]; - #ifndef UNDER_CE - WCHAR workCurDir[MAX_PATH + 32]; - #endif - size_t pathLen; - DWORD winRes; - const wchar_t *cmdLineParams; - const char *errorMessage = NULL; - BoolInt useShellExecute = True; - DWORD exitCode = 0; - - LoadSecurityDlls(); - - #ifdef _CONSOLE - SetConsoleCtrlHandler(HandlerRoutine, TRUE); - #else - UNUSED_VAR(hInstance); - UNUSED_VAR(hPrevInstance); - UNUSED_VAR(lpCmdLine); - UNUSED_VAR(nCmdShow); - #endif - - CrcGenerateTable(); - - allocImp.Alloc = SzAlloc; - allocImp.Free = SzFree; - - allocTempImp.Alloc = SzAllocTemp; - allocTempImp.Free = SzFreeTemp; - - FileInStream_CreateVTable(&archiveStream); - LookToRead2_CreateVTable(&lookStream, False); - lookStream.buf = NULL; - - winRes = GetModuleFileNameW(NULL, sfxPath, MAX_PATH); - if (winRes == 0 || winRes > MAX_PATH) - return 1; - { - cmdLineParams = GetCommandLineW(); - #ifndef UNDER_CE - { - BoolInt quoteMode = False; - for (;; cmdLineParams++) - { - wchar_t c = *cmdLineParams; - if (c == L'\"') - quoteMode = !quoteMode; - else if (c == 0 || (c == L' ' && !quoteMode)) - break; - } - } - #endif - } - - { - unsigned i; - DWORD d; - winRes = GetTempPathW(MAX_PATH, path); - if (winRes == 0 || winRes > MAX_PATH) - return 1; - pathLen = wcslen(path); - d = (GetTickCount() << 12) ^ (GetCurrentThreadId() << 14) ^ GetCurrentProcessId(); - - for (i = 0;; i++, d += GetTickCount()) - { - if (i >= 100) - { - res = SZ_ERROR_FAIL; - break; - } - wcscpy(path + pathLen, L"7z"); - - { - wchar_t *s = path + wcslen(path); - UInt32 value = d; - unsigned k; - for (k = 0; k < 8; k++) - { - unsigned t = value & 0xF; - value >>= 4; - s[7 - k] = (wchar_t)((t < 10) ? ('0' + t) : ('A' + (t - 10))); - } - s[k] = '\0'; - } - - if (DoesFileOrDirExist(path)) - continue; - if (CreateDirectoryW(path, NULL)) - { - wcscat(path, WSTRING_PATH_SEPARATOR); - pathLen = wcslen(path); - break; - } - if (GetLastError() != ERROR_ALREADY_EXISTS) - { - res = SZ_ERROR_FAIL; - break; - } - } - - #ifndef UNDER_CE - wcscpy(workCurDir, path); - #endif - if (res != SZ_OK) - errorMessage = "Can't create temp folder"; - } - - if (res != SZ_OK) - { - if (!errorMessage) - errorMessage = "Error"; - PrintErrorMessage(errorMessage); - return 1; - } - - if (InFile_OpenW(&archiveStream.file, sfxPath) != 0) - { - errorMessage = "can not open input file"; - res = SZ_ERROR_FAIL; - } - else - { - UInt64 pos = 0; - if (!FindSignature(&archiveStream.file, &pos)) - res = SZ_ERROR_FAIL; - else if (File_Seek(&archiveStream.file, (Int64 *)&pos, SZ_SEEK_SET) != 0) - res = SZ_ERROR_FAIL; - if (res != 0) - errorMessage = "Can't find 7z archive"; - } - - if (res == SZ_OK) - { - lookStream.buf = (Byte *)ISzAlloc_Alloc(&allocImp, kInputBufSize); - if (!lookStream.buf) - res = SZ_ERROR_MEM; - else - { - lookStream.bufSize = kInputBufSize; - lookStream.realStream = &archiveStream.vt; - LookToRead2_Init(&lookStream); - } - } - - SzArEx_Init(&db); - - if (res == SZ_OK) - { - res = SzArEx_Open(&db, &lookStream.vt, &allocImp, &allocTempImp); - } - - if (res == SZ_OK) - { - UInt32 executeFileIndex = (UInt32)(Int32)-1; - UInt32 minPrice = 1 << 30; - UInt32 i; - UInt32 blockIndex = 0xFFFFFFFF; /* it can have any value before first call (if outBuffer = 0) */ - Byte *outBuffer = 0; /* it must be 0 before first call for each new archive. */ - size_t outBufferSize = 0; /* it can have any value before first call (if outBuffer = 0) */ - - for (i = 0; i < db.NumFiles; i++) - { - size_t offset = 0; - size_t outSizeProcessed = 0; - WCHAR *temp; - - if (SzArEx_GetFileNameUtf16(&db, i, NULL) >= MAX_PATH) - { - res = SZ_ERROR_FAIL; - break; - } - - temp = path + pathLen; - - SzArEx_GetFileNameUtf16(&db, i, (UInt16 *)temp); - { - res = SzArEx_Extract(&db, &lookStream.vt, i, - &blockIndex, &outBuffer, &outBufferSize, - &offset, &outSizeProcessed, - &allocImp, &allocTempImp); - if (res != SZ_OK) - break; - } - { - CSzFile outFile; - size_t processedSize; - size_t j; - size_t nameStartPos = 0; - for (j = 0; temp[j] != 0; j++) - { - if (temp[j] == '/') - { - temp[j] = 0; - MyCreateDir(path); - temp[j] = CHAR_PATH_SEPARATOR; - nameStartPos = j + 1; - } - } - - if (SzArEx_IsDir(&db, i)) - { - MyCreateDir(path); - continue; - } - else - { - unsigned extLen; - const WCHAR *name = temp + nameStartPos; - unsigned len = (unsigned)wcslen(name); - unsigned nameLen = FindExt(temp + nameStartPos, &extLen); - unsigned extPrice = FindItem(kExts, sizeof(kExts) / sizeof(kExts[0]), name + len - extLen, extLen); - unsigned namePrice = FindItem(kNames, sizeof(kNames) / sizeof(kNames[0]), name, nameLen); - - unsigned price = namePrice + extPrice * 64 + (nameStartPos == 0 ? 0 : (1 << 12)); - if (minPrice > price) - { - minPrice = price; - executeFileIndex = i; - useShellExecute = (extPrice != k_EXE_ExtIndex); - } - - if (DoesFileOrDirExist(path)) - { - errorMessage = "Duplicate file"; - res = SZ_ERROR_FAIL; - break; - } - if (OutFile_OpenW(&outFile, path)) - { - errorMessage = "Can't open output file"; - res = SZ_ERROR_FAIL; - break; - } - } - - processedSize = outSizeProcessed; - if (File_Write(&outFile, outBuffer + offset, &processedSize) != 0 || processedSize != outSizeProcessed) - { - errorMessage = "Can't write output file"; - res = SZ_ERROR_FAIL; - } - - #ifdef USE_WINDOWS_FILE - if (SzBitWithVals_Check(&db.MTime, i)) - { - const CNtfsFileTime *t = db.MTime.Vals + i; - FILETIME mTime; - mTime.dwLowDateTime = t->Low; - mTime.dwHighDateTime = t->High; - SetFileTime(outFile.handle, NULL, NULL, &mTime); - } - #endif - - { - SRes res2 = File_Close(&outFile); - if (res != SZ_OK) - break; - if (res2 != SZ_OK) - { - res = res2; - break; - } - } - #ifdef USE_WINDOWS_FILE - if (SzBitWithVals_Check(&db.Attribs, i)) - SetFileAttributesW(path, db.Attribs.Vals[i]); - #endif - } - } - - if (res == SZ_OK) - { - if (executeFileIndex == (UInt32)(Int32)-1) - { - errorMessage = "There is no file to execute"; - res = SZ_ERROR_FAIL; - } - else - { - WCHAR *temp = path + pathLen; - UInt32 j; - SzArEx_GetFileNameUtf16(&db, executeFileIndex, (UInt16 *)temp); - for (j = 0; temp[j] != 0; j++) - if (temp[j] == '/') - temp[j] = CHAR_PATH_SEPARATOR; - } - } - ISzAlloc_Free(&allocImp, outBuffer); - } - - SzArEx_Free(&db, &allocImp); - - ISzAlloc_Free(&allocImp, lookStream.buf); - - File_Close(&archiveStream.file); - - if (res == SZ_OK) - { - HANDLE hProcess = 0; - - #ifndef UNDER_CE - WCHAR oldCurDir[MAX_PATH + 2]; - oldCurDir[0] = 0; - { - DWORD needLen = GetCurrentDirectory(MAX_PATH + 1, oldCurDir); - if (needLen == 0 || needLen > MAX_PATH) - oldCurDir[0] = 0; - SetCurrentDirectory(workCurDir); - } - #endif - - if (useShellExecute) - { - SHELLEXECUTEINFO ei; - UINT32 executeRes; - BOOL success; - - memset(&ei, 0, sizeof(ei)); - ei.cbSize = sizeof(ei); - ei.lpFile = path; - ei.fMask = SEE_MASK_NOCLOSEPROCESS - #ifndef UNDER_CE - | SEE_MASK_FLAG_DDEWAIT - #endif - /* | SEE_MASK_NO_CONSOLE */ - ; - if (wcslen(cmdLineParams) != 0) - ei.lpParameters = cmdLineParams; - ei.nShow = SW_SHOWNORMAL; /* SW_HIDE; */ - success = ShellExecuteEx(&ei); - executeRes = (UINT32)(UINT_PTR)ei.hInstApp; - if (!success || (executeRes <= 32 && executeRes != 0)) /* executeRes = 0 in Windows CE */ - res = SZ_ERROR_FAIL; - else - hProcess = ei.hProcess; - } - else - { - STARTUPINFOW si; - PROCESS_INFORMATION pi; - WCHAR cmdLine[MAX_PATH * 3]; - - wcscpy(cmdLine, path); - wcscat(cmdLine, cmdLineParams); - memset(&si, 0, sizeof(si)); - si.cb = sizeof(si); - if (CreateProcessW(NULL, cmdLine, NULL, NULL, FALSE, 0, NULL, NULL, &si, &pi) == 0) - res = SZ_ERROR_FAIL; - else - { - CloseHandle(pi.hThread); - hProcess = pi.hProcess; - } - } - - if (hProcess != 0) - { - WaitForSingleObject(hProcess, INFINITE); - if (!GetExitCodeProcess(hProcess, &exitCode)) - exitCode = 1; - CloseHandle(hProcess); - } - - #ifndef UNDER_CE - SetCurrentDirectory(oldCurDir); - #endif - } - - path[pathLen] = L'\0'; - RemoveDirWithSubItems(path); - - if (res == SZ_OK) - return (int)exitCode; - - { - if (res == SZ_ERROR_UNSUPPORTED) - errorMessage = "Decoder doesn't support this archive"; - else if (res == SZ_ERROR_MEM) - errorMessage = "Can't allocate required memory"; - else if (res == SZ_ERROR_CRC) - errorMessage = "CRC error"; - else - { - if (!errorMessage) - errorMessage = "ERROR"; - } - - if (errorMessage) - PrintErrorMessage(errorMessage); - } - return 1; -} +/* SfxSetup.c - 7z SFX Setup +2019-02-02 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#ifndef UNICODE +#define UNICODE +#endif + +#ifndef _UNICODE +#define _UNICODE +#endif + +#ifdef _CONSOLE +#include +#endif + +#include "../../7z.h" +#include "../../7zAlloc.h" +#include "../../7zCrc.h" +#include "../../7zFile.h" +#include "../../CpuArch.h" +#include "../../DllSecur.h" + +#define k_EXE_ExtIndex 2 + +#define kInputBufSize ((size_t)1 << 18) + +static const char * const kExts[] = +{ + "bat" + , "cmd" + , "exe" + , "inf" + , "msi" + #ifdef UNDER_CE + , "cab" + #endif + , "html" + , "htm" +}; + +static const char * const kNames[] = +{ + "setup" + , "install" + , "run" + , "start" +}; + +static unsigned FindExt(const wchar_t *s, unsigned *extLen) +{ + unsigned len = (unsigned)wcslen(s); + unsigned i; + for (i = len; i > 0; i--) + { + if (s[i - 1] == '.') + { + *extLen = len - i; + return i - 1; + } + } + *extLen = 0; + return len; +} + +#define MAKE_CHAR_UPPER(c) ((((c) >= 'a' && (c) <= 'z') ? (c) -= 0x20 : (c))) + +static unsigned FindItem(const char * const *items, unsigned num, const wchar_t *s, unsigned len) +{ + unsigned i; + for (i = 0; i < num; i++) + { + const char *item = items[i]; + unsigned itemLen = (unsigned)strlen(item); + unsigned j; + if (len != itemLen) + continue; + for (j = 0; j < len; j++) + { + unsigned c = (Byte)item[j]; + if (c != s[j] && MAKE_CHAR_UPPER(c) != s[j]) + break; + } + if (j == len) + return i; + } + return i; +} + +#ifdef _CONSOLE +static BOOL WINAPI HandlerRoutine(DWORD ctrlType) +{ + UNUSED_VAR(ctrlType); + return TRUE; +} +#endif + +static void PrintErrorMessage(const char *message) +{ + #ifdef _CONSOLE + printf("\n7-Zip Error: %s\n", message); + #else + #ifdef UNDER_CE + WCHAR messageW[256 + 4]; + unsigned i; + for (i = 0; i < 256 && message[i] != 0; i++) + messageW[i] = message[i]; + messageW[i] = 0; + MessageBoxW(0, messageW, L"7-Zip Error", MB_ICONERROR); + #else + MessageBoxA(0, message, "7-Zip Error", MB_ICONERROR); + #endif + #endif +} + +static WRes MyCreateDir(const WCHAR *name) +{ + return CreateDirectoryW(name, NULL) ? 0 : GetLastError(); +} + +#ifdef UNDER_CE +#define kBufferSize (1 << 13) +#else +#define kBufferSize (1 << 15) +#endif + +#define kSignatureSearchLimit (1 << 22) + +static BoolInt FindSignature(CSzFile *stream, UInt64 *resPos) +{ + Byte buf[kBufferSize]; + size_t numPrevBytes = 0; + *resPos = 0; + for (;;) + { + size_t processed, pos; + if (*resPos > kSignatureSearchLimit) + return False; + processed = kBufferSize - numPrevBytes; + if (File_Read(stream, buf + numPrevBytes, &processed) != 0) + return False; + processed += numPrevBytes; + if (processed < k7zStartHeaderSize || + (processed == k7zStartHeaderSize && numPrevBytes != 0)) + return False; + processed -= k7zStartHeaderSize; + for (pos = 0; pos <= processed; pos++) + { + for (; pos <= processed && buf[pos] != '7'; pos++); + if (pos > processed) + break; + if (memcmp(buf + pos, k7zSignature, k7zSignatureSize) == 0) + if (CrcCalc(buf + pos + 12, 20) == GetUi32(buf + pos + 8)) + { + *resPos += pos; + return True; + } + } + *resPos += processed; + numPrevBytes = k7zStartHeaderSize; + memmove(buf, buf + processed, k7zStartHeaderSize); + } +} + +static BoolInt DoesFileOrDirExist(const WCHAR *path) +{ + WIN32_FIND_DATAW fd; + HANDLE handle; + handle = FindFirstFileW(path, &fd); + if (handle == INVALID_HANDLE_VALUE) + return False; + FindClose(handle); + return True; +} + +static WRes RemoveDirWithSubItems(WCHAR *path) +{ + WIN32_FIND_DATAW fd; + HANDLE handle; + WRes res = 0; + size_t len = wcslen(path); + wcscpy(path + len, L"*"); + handle = FindFirstFileW(path, &fd); + path[len] = L'\0'; + if (handle == INVALID_HANDLE_VALUE) + return GetLastError(); + + for (;;) + { + if (wcscmp(fd.cFileName, L".") != 0 && + wcscmp(fd.cFileName, L"..") != 0) + { + wcscpy(path + len, fd.cFileName); + if ((fd.dwFileAttributes & FILE_ATTRIBUTE_DIRECTORY) != 0) + { + wcscat(path, WSTRING_PATH_SEPARATOR); + res = RemoveDirWithSubItems(path); + } + else + { + SetFileAttributesW(path, 0); + if (DeleteFileW(path) == 0) + res = GetLastError(); + } + + if (res != 0) + break; + } + + if (!FindNextFileW(handle, &fd)) + { + res = GetLastError(); + if (res == ERROR_NO_MORE_FILES) + res = 0; + break; + } + } + + path[len] = L'\0'; + FindClose(handle); + if (res == 0) + { + if (!RemoveDirectoryW(path)) + res = GetLastError(); + } + return res; +} + +#ifdef _CONSOLE +int MY_CDECL main() +#else +int APIENTRY WinMain(HINSTANCE hInstance, HINSTANCE hPrevInstance, + #ifdef UNDER_CE + LPWSTR + #else + LPSTR + #endif + lpCmdLine, int nCmdShow) +#endif +{ + CFileInStream archiveStream; + CLookToRead2 lookStream; + CSzArEx db; + SRes res = SZ_OK; + ISzAlloc allocImp; + ISzAlloc allocTempImp; + WCHAR sfxPath[MAX_PATH + 2]; + WCHAR path[MAX_PATH * 3 + 2]; + #ifndef UNDER_CE + WCHAR workCurDir[MAX_PATH + 32]; + #endif + size_t pathLen; + DWORD winRes; + const wchar_t *cmdLineParams; + const char *errorMessage = NULL; + BoolInt useShellExecute = True; + DWORD exitCode = 0; + + LoadSecurityDlls(); + + #ifdef _CONSOLE + SetConsoleCtrlHandler(HandlerRoutine, TRUE); + #else + UNUSED_VAR(hInstance); + UNUSED_VAR(hPrevInstance); + UNUSED_VAR(lpCmdLine); + UNUSED_VAR(nCmdShow); + #endif + + CrcGenerateTable(); + + allocImp.Alloc = SzAlloc; + allocImp.Free = SzFree; + + allocTempImp.Alloc = SzAllocTemp; + allocTempImp.Free = SzFreeTemp; + + FileInStream_CreateVTable(&archiveStream); + LookToRead2_CreateVTable(&lookStream, False); + lookStream.buf = NULL; + + winRes = GetModuleFileNameW(NULL, sfxPath, MAX_PATH); + if (winRes == 0 || winRes > MAX_PATH) + return 1; + { + cmdLineParams = GetCommandLineW(); + #ifndef UNDER_CE + { + BoolInt quoteMode = False; + for (;; cmdLineParams++) + { + wchar_t c = *cmdLineParams; + if (c == L'\"') + quoteMode = !quoteMode; + else if (c == 0 || (c == L' ' && !quoteMode)) + break; + } + } + #endif + } + + { + unsigned i; + DWORD d; + winRes = GetTempPathW(MAX_PATH, path); + if (winRes == 0 || winRes > MAX_PATH) + return 1; + pathLen = wcslen(path); + d = (GetTickCount() << 12) ^ (GetCurrentThreadId() << 14) ^ GetCurrentProcessId(); + + for (i = 0;; i++, d += GetTickCount()) + { + if (i >= 100) + { + res = SZ_ERROR_FAIL; + break; + } + wcscpy(path + pathLen, L"7z"); + + { + wchar_t *s = path + wcslen(path); + UInt32 value = d; + unsigned k; + for (k = 0; k < 8; k++) + { + unsigned t = value & 0xF; + value >>= 4; + s[7 - k] = (wchar_t)((t < 10) ? ('0' + t) : ('A' + (t - 10))); + } + s[k] = '\0'; + } + + if (DoesFileOrDirExist(path)) + continue; + if (CreateDirectoryW(path, NULL)) + { + wcscat(path, WSTRING_PATH_SEPARATOR); + pathLen = wcslen(path); + break; + } + if (GetLastError() != ERROR_ALREADY_EXISTS) + { + res = SZ_ERROR_FAIL; + break; + } + } + + #ifndef UNDER_CE + wcscpy(workCurDir, path); + #endif + if (res != SZ_OK) + errorMessage = "Can't create temp folder"; + } + + if (res != SZ_OK) + { + if (!errorMessage) + errorMessage = "Error"; + PrintErrorMessage(errorMessage); + return 1; + } + + if (InFile_OpenW(&archiveStream.file, sfxPath) != 0) + { + errorMessage = "can not open input file"; + res = SZ_ERROR_FAIL; + } + else + { + UInt64 pos = 0; + if (!FindSignature(&archiveStream.file, &pos)) + res = SZ_ERROR_FAIL; + else if (File_Seek(&archiveStream.file, (Int64 *)&pos, SZ_SEEK_SET) != 0) + res = SZ_ERROR_FAIL; + if (res != 0) + errorMessage = "Can't find 7z archive"; + } + + if (res == SZ_OK) + { + lookStream.buf = (Byte *)ISzAlloc_Alloc(&allocImp, kInputBufSize); + if (!lookStream.buf) + res = SZ_ERROR_MEM; + else + { + lookStream.bufSize = kInputBufSize; + lookStream.realStream = &archiveStream.vt; + LookToRead2_Init(&lookStream); + } + } + + SzArEx_Init(&db); + + if (res == SZ_OK) + { + res = SzArEx_Open(&db, &lookStream.vt, &allocImp, &allocTempImp); + } + + if (res == SZ_OK) + { + UInt32 executeFileIndex = (UInt32)(Int32)-1; + UInt32 minPrice = 1 << 30; + UInt32 i; + UInt32 blockIndex = 0xFFFFFFFF; /* it can have any value before first call (if outBuffer = 0) */ + Byte *outBuffer = 0; /* it must be 0 before first call for each new archive. */ + size_t outBufferSize = 0; /* it can have any value before first call (if outBuffer = 0) */ + + for (i = 0; i < db.NumFiles; i++) + { + size_t offset = 0; + size_t outSizeProcessed = 0; + WCHAR *temp; + + if (SzArEx_GetFileNameUtf16(&db, i, NULL) >= MAX_PATH) + { + res = SZ_ERROR_FAIL; + break; + } + + temp = path + pathLen; + + SzArEx_GetFileNameUtf16(&db, i, (UInt16 *)temp); + { + res = SzArEx_Extract(&db, &lookStream.vt, i, + &blockIndex, &outBuffer, &outBufferSize, + &offset, &outSizeProcessed, + &allocImp, &allocTempImp); + if (res != SZ_OK) + break; + } + { + CSzFile outFile; + size_t processedSize; + size_t j; + size_t nameStartPos = 0; + for (j = 0; temp[j] != 0; j++) + { + if (temp[j] == '/') + { + temp[j] = 0; + MyCreateDir(path); + temp[j] = CHAR_PATH_SEPARATOR; + nameStartPos = j + 1; + } + } + + if (SzArEx_IsDir(&db, i)) + { + MyCreateDir(path); + continue; + } + else + { + unsigned extLen; + const WCHAR *name = temp + nameStartPos; + unsigned len = (unsigned)wcslen(name); + unsigned nameLen = FindExt(temp + nameStartPos, &extLen); + unsigned extPrice = FindItem(kExts, sizeof(kExts) / sizeof(kExts[0]), name + len - extLen, extLen); + unsigned namePrice = FindItem(kNames, sizeof(kNames) / sizeof(kNames[0]), name, nameLen); + + unsigned price = namePrice + extPrice * 64 + (nameStartPos == 0 ? 0 : (1 << 12)); + if (minPrice > price) + { + minPrice = price; + executeFileIndex = i; + useShellExecute = (extPrice != k_EXE_ExtIndex); + } + + if (DoesFileOrDirExist(path)) + { + errorMessage = "Duplicate file"; + res = SZ_ERROR_FAIL; + break; + } + if (OutFile_OpenW(&outFile, path)) + { + errorMessage = "Can't open output file"; + res = SZ_ERROR_FAIL; + break; + } + } + + processedSize = outSizeProcessed; + if (File_Write(&outFile, outBuffer + offset, &processedSize) != 0 || processedSize != outSizeProcessed) + { + errorMessage = "Can't write output file"; + res = SZ_ERROR_FAIL; + } + + #ifdef USE_WINDOWS_FILE + if (SzBitWithVals_Check(&db.MTime, i)) + { + const CNtfsFileTime *t = db.MTime.Vals + i; + FILETIME mTime; + mTime.dwLowDateTime = t->Low; + mTime.dwHighDateTime = t->High; + SetFileTime(outFile.handle, NULL, NULL, &mTime); + } + #endif + + { + SRes res2 = File_Close(&outFile); + if (res != SZ_OK) + break; + if (res2 != SZ_OK) + { + res = res2; + break; + } + } + #ifdef USE_WINDOWS_FILE + if (SzBitWithVals_Check(&db.Attribs, i)) + SetFileAttributesW(path, db.Attribs.Vals[i]); + #endif + } + } + + if (res == SZ_OK) + { + if (executeFileIndex == (UInt32)(Int32)-1) + { + errorMessage = "There is no file to execute"; + res = SZ_ERROR_FAIL; + } + else + { + WCHAR *temp = path + pathLen; + UInt32 j; + SzArEx_GetFileNameUtf16(&db, executeFileIndex, (UInt16 *)temp); + for (j = 0; temp[j] != 0; j++) + if (temp[j] == '/') + temp[j] = CHAR_PATH_SEPARATOR; + } + } + ISzAlloc_Free(&allocImp, outBuffer); + } + + SzArEx_Free(&db, &allocImp); + + ISzAlloc_Free(&allocImp, lookStream.buf); + + File_Close(&archiveStream.file); + + if (res == SZ_OK) + { + HANDLE hProcess = 0; + + #ifndef UNDER_CE + WCHAR oldCurDir[MAX_PATH + 2]; + oldCurDir[0] = 0; + { + DWORD needLen = GetCurrentDirectory(MAX_PATH + 1, oldCurDir); + if (needLen == 0 || needLen > MAX_PATH) + oldCurDir[0] = 0; + SetCurrentDirectory(workCurDir); + } + #endif + + if (useShellExecute) + { + SHELLEXECUTEINFO ei; + UINT32 executeRes; + BOOL success; + + memset(&ei, 0, sizeof(ei)); + ei.cbSize = sizeof(ei); + ei.lpFile = path; + ei.fMask = SEE_MASK_NOCLOSEPROCESS + #ifndef UNDER_CE + | SEE_MASK_FLAG_DDEWAIT + #endif + /* | SEE_MASK_NO_CONSOLE */ + ; + if (wcslen(cmdLineParams) != 0) + ei.lpParameters = cmdLineParams; + ei.nShow = SW_SHOWNORMAL; /* SW_HIDE; */ + success = ShellExecuteEx(&ei); + executeRes = (UINT32)(UINT_PTR)ei.hInstApp; + if (!success || (executeRes <= 32 && executeRes != 0)) /* executeRes = 0 in Windows CE */ + res = SZ_ERROR_FAIL; + else + hProcess = ei.hProcess; + } + else + { + STARTUPINFOW si; + PROCESS_INFORMATION pi; + WCHAR cmdLine[MAX_PATH * 3]; + + wcscpy(cmdLine, path); + wcscat(cmdLine, cmdLineParams); + memset(&si, 0, sizeof(si)); + si.cb = sizeof(si); + if (CreateProcessW(NULL, cmdLine, NULL, NULL, FALSE, 0, NULL, NULL, &si, &pi) == 0) + res = SZ_ERROR_FAIL; + else + { + CloseHandle(pi.hThread); + hProcess = pi.hProcess; + } + } + + if (hProcess != 0) + { + WaitForSingleObject(hProcess, INFINITE); + if (!GetExitCodeProcess(hProcess, &exitCode)) + exitCode = 1; + CloseHandle(hProcess); + } + + #ifndef UNDER_CE + SetCurrentDirectory(oldCurDir); + #endif + } + + path[pathLen] = L'\0'; + RemoveDirWithSubItems(path); + + if (res == SZ_OK) + return (int)exitCode; + + { + if (res == SZ_ERROR_UNSUPPORTED) + errorMessage = "Decoder doesn't support this archive"; + else if (res == SZ_ERROR_MEM) + errorMessage = "Can't allocate required memory"; + else if (res == SZ_ERROR_CRC) + errorMessage = "CRC error"; + else + { + if (!errorMessage) + errorMessage = "ERROR"; + } + + if (errorMessage) + PrintErrorMessage(errorMessage); + } + return 1; +} diff --git a/deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.dsp b/deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.dsp index be9de6dec..60439a6f7 100644 --- a/deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.dsp +++ b/deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.dsp @@ -1,231 +1,231 @@ -# Microsoft Developer Studio Project File - Name="SfxSetup" - Package Owner=<4> -# Microsoft Developer Studio Generated Build File, Format Version 6.00 -# ** DO NOT EDIT ** - -# TARGTYPE "Win32 (x86) Application" 0x0101 - -CFG=SfxSetup - Win32 Debug -!MESSAGE This is not a valid makefile. To build this project using NMAKE, -!MESSAGE use the Export Makefile command and run -!MESSAGE -!MESSAGE NMAKE /f "SfxSetup.mak". -!MESSAGE -!MESSAGE You can specify a configuration when running NMAKE -!MESSAGE by defining the macro CFG on the command line. For example: -!MESSAGE -!MESSAGE NMAKE /f "SfxSetup.mak" CFG="SfxSetup - Win32 Debug" -!MESSAGE -!MESSAGE Possible choices for configuration are: -!MESSAGE -!MESSAGE "SfxSetup - Win32 Release" (based on "Win32 (x86) Application") -!MESSAGE "SfxSetup - Win32 Debug" (based on "Win32 (x86) Application") -!MESSAGE - -# Begin Project -# PROP AllowPerConfigDependencies 0 -# PROP Scc_ProjName "" -# PROP Scc_LocalPath "" -CPP=cl.exe -MTL=midl.exe -RSC=rc.exe - -!IF "$(CFG)" == "SfxSetup - Win32 Release" - -# PROP BASE Use_MFC 0 -# PROP BASE Use_Debug_Libraries 0 -# PROP BASE Output_Dir "Release" -# PROP BASE Intermediate_Dir "Release" -# PROP BASE Target_Dir "" -# PROP Use_MFC 0 -# PROP Use_Debug_Libraries 0 -# PROP Output_Dir "Release" -# PROP Intermediate_Dir "Release" -# PROP Target_Dir "" -# ADD BASE CPP /nologo /W3 /GX /O2 /D "WIN32" /D "NDEBUG" /D "_WINDOWS" /D "_MBCS" /YX /FD /c -# ADD CPP /nologo /W4 /WX /GX /O2 /D "NDEBUG" /D "WIN32" /D "_WINDOWS" /D "_UNICODE" /D "UNICODE" /Yu"Precomp.h" /FD /c -# ADD BASE MTL /nologo /D "NDEBUG" /mktyplib203 /win32 -# ADD MTL /nologo /D "NDEBUG" /mktyplib203 /win32 -# ADD BASE RSC /l 0x419 /d "NDEBUG" -# ADD RSC /l 0x419 /d "NDEBUG" -BSC32=bscmake.exe -# ADD BASE BSC32 /nologo -# ADD BSC32 /nologo -LINK32=link.exe -# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /machine:I386 -# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /machine:I386 - -!ELSEIF "$(CFG)" == "SfxSetup - Win32 Debug" - -# PROP BASE Use_MFC 0 -# PROP BASE Use_Debug_Libraries 1 -# PROP BASE Output_Dir "Debug" -# PROP BASE Intermediate_Dir "Debug" -# PROP BASE Target_Dir "" -# PROP Use_MFC 0 -# PROP Use_Debug_Libraries 1 -# PROP Output_Dir "Debug" -# PROP Intermediate_Dir "Debug" -# PROP Ignore_Export_Lib 0 -# PROP Target_Dir "" -# ADD BASE CPP /nologo /W3 /Gm /GX /ZI /Od /D "WIN32" /D "_DEBUG" /D "_WINDOWS" /D "_MBCS" /YX /FD /GZ /c -# ADD CPP /nologo /W4 /WX /Gm /GX /ZI /Od /D "_DEBUG" /D "WIN32" /D "_WINDOWS" /D "_UNICODE" /D "UNICODE" /Yu"Precomp.h" /FD /GZ /c -# ADD BASE MTL /nologo /D "_DEBUG" /mktyplib203 /win32 -# ADD MTL /nologo /D "_DEBUG" /mktyplib203 /win32 -# ADD BASE RSC /l 0x419 /d "_DEBUG" -# ADD RSC /l 0x419 /d "_DEBUG" -BSC32=bscmake.exe -# ADD BASE BSC32 /nologo -# ADD BSC32 /nologo -LINK32=link.exe -# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /debug /machine:I386 /pdbtype:sept -# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /debug /machine:I386 /pdbtype:sept - -!ENDIF - -# Begin Target - -# Name "SfxSetup - Win32 Release" -# Name "SfxSetup - Win32 Debug" -# Begin Group "Common" - -# PROP Default_Filter "" -# Begin Source File - -SOURCE=..\..\7z.h -# End Source File -# Begin Source File - -SOURCE=..\..\7zAlloc.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zAlloc.h -# End Source File -# Begin Source File - -SOURCE=..\..\7zArcIn.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zBuf.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zBuf.h -# End Source File -# Begin Source File - -SOURCE=..\..\7zCrc.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zCrc.h -# End Source File -# Begin Source File - -SOURCE=..\..\7zCrcOpt.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zDec.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zFile.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zFile.h -# End Source File -# Begin Source File - -SOURCE=..\..\7zStream.c -# End Source File -# Begin Source File - -SOURCE=..\..\7zTypes.h -# End Source File -# Begin Source File - -SOURCE=..\..\Bcj2.c -# End Source File -# Begin Source File - -SOURCE=..\..\Bcj2.h -# End Source File -# Begin Source File - -SOURCE=..\..\Bra.c -# End Source File -# Begin Source File - -SOURCE=..\..\Bra.h -# End Source File -# Begin Source File - -SOURCE=..\..\Bra86.c -# End Source File -# Begin Source File - -SOURCE=..\..\BraIA64.c -# End Source File -# Begin Source File - -SOURCE=..\..\CpuArch.c -# End Source File -# Begin Source File - -SOURCE=..\..\CpuArch.h -# End Source File -# Begin Source File - -SOURCE=..\..\Delta.c -# End Source File -# Begin Source File - -SOURCE=..\..\Delta.h -# End Source File -# Begin Source File - -SOURCE=..\..\DllSecur.c -# End Source File -# Begin Source File - -SOURCE=..\..\DllSecur.h -# End Source File -# Begin Source File - -SOURCE=..\..\Lzma2Dec.c -# End Source File -# Begin Source File - -SOURCE=..\..\Lzma2Dec.h -# End Source File -# Begin Source File - -SOURCE=..\..\LzmaDec.c -# End Source File -# Begin Source File - -SOURCE=..\..\LzmaDec.h -# End Source File -# End Group -# Begin Group "Spec" - -# PROP Default_Filter "" -# Begin Source File - -SOURCE=.\Precomp.c -# ADD CPP /Yc"Precomp.h" -# End Source File -# Begin Source File - -SOURCE=.\Precomp.h -# End Source File -# End Group -# Begin Source File - -SOURCE=.\SfxSetup.c -# End Source File -# End Target -# End Project +# Microsoft Developer Studio Project File - Name="SfxSetup" - Package Owner=<4> +# Microsoft Developer Studio Generated Build File, Format Version 6.00 +# ** DO NOT EDIT ** + +# TARGTYPE "Win32 (x86) Application" 0x0101 + +CFG=SfxSetup - Win32 Debug +!MESSAGE This is not a valid makefile. To build this project using NMAKE, +!MESSAGE use the Export Makefile command and run +!MESSAGE +!MESSAGE NMAKE /f "SfxSetup.mak". +!MESSAGE +!MESSAGE You can specify a configuration when running NMAKE +!MESSAGE by defining the macro CFG on the command line. For example: +!MESSAGE +!MESSAGE NMAKE /f "SfxSetup.mak" CFG="SfxSetup - Win32 Debug" +!MESSAGE +!MESSAGE Possible choices for configuration are: +!MESSAGE +!MESSAGE "SfxSetup - Win32 Release" (based on "Win32 (x86) Application") +!MESSAGE "SfxSetup - Win32 Debug" (based on "Win32 (x86) Application") +!MESSAGE + +# Begin Project +# PROP AllowPerConfigDependencies 0 +# PROP Scc_ProjName "" +# PROP Scc_LocalPath "" +CPP=cl.exe +MTL=midl.exe +RSC=rc.exe + +!IF "$(CFG)" == "SfxSetup - Win32 Release" + +# PROP BASE Use_MFC 0 +# PROP BASE Use_Debug_Libraries 0 +# PROP BASE Output_Dir "Release" +# PROP BASE Intermediate_Dir "Release" +# PROP BASE Target_Dir "" +# PROP Use_MFC 0 +# PROP Use_Debug_Libraries 0 +# PROP Output_Dir "Release" +# PROP Intermediate_Dir "Release" +# PROP Target_Dir "" +# ADD BASE CPP /nologo /W3 /GX /O2 /D "WIN32" /D "NDEBUG" /D "_WINDOWS" /D "_MBCS" /YX /FD /c +# ADD CPP /nologo /W4 /WX /GX /O2 /D "NDEBUG" /D "WIN32" /D "_WINDOWS" /D "_UNICODE" /D "UNICODE" /Yu"Precomp.h" /FD /c +# ADD BASE MTL /nologo /D "NDEBUG" /mktyplib203 /win32 +# ADD MTL /nologo /D "NDEBUG" /mktyplib203 /win32 +# ADD BASE RSC /l 0x419 /d "NDEBUG" +# ADD RSC /l 0x419 /d "NDEBUG" +BSC32=bscmake.exe +# ADD BASE BSC32 /nologo +# ADD BSC32 /nologo +LINK32=link.exe +# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /machine:I386 +# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /machine:I386 + +!ELSEIF "$(CFG)" == "SfxSetup - Win32 Debug" + +# PROP BASE Use_MFC 0 +# PROP BASE Use_Debug_Libraries 1 +# PROP BASE Output_Dir "Debug" +# PROP BASE Intermediate_Dir "Debug" +# PROP BASE Target_Dir "" +# PROP Use_MFC 0 +# PROP Use_Debug_Libraries 1 +# PROP Output_Dir "Debug" +# PROP Intermediate_Dir "Debug" +# PROP Ignore_Export_Lib 0 +# PROP Target_Dir "" +# ADD BASE CPP /nologo /W3 /Gm /GX /ZI /Od /D "WIN32" /D "_DEBUG" /D "_WINDOWS" /D "_MBCS" /YX /FD /GZ /c +# ADD CPP /nologo /W4 /WX /Gm /GX /ZI /Od /D "_DEBUG" /D "WIN32" /D "_WINDOWS" /D "_UNICODE" /D "UNICODE" /Yu"Precomp.h" /FD /GZ /c +# ADD BASE MTL /nologo /D "_DEBUG" /mktyplib203 /win32 +# ADD MTL /nologo /D "_DEBUG" /mktyplib203 /win32 +# ADD BASE RSC /l 0x419 /d "_DEBUG" +# ADD RSC /l 0x419 /d "_DEBUG" +BSC32=bscmake.exe +# ADD BASE BSC32 /nologo +# ADD BSC32 /nologo +LINK32=link.exe +# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /debug /machine:I386 /pdbtype:sept +# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /debug /machine:I386 /pdbtype:sept + +!ENDIF + +# Begin Target + +# Name "SfxSetup - Win32 Release" +# Name "SfxSetup - Win32 Debug" +# Begin Group "Common" + +# PROP Default_Filter "" +# Begin Source File + +SOURCE=..\..\7z.h +# End Source File +# Begin Source File + +SOURCE=..\..\7zAlloc.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zAlloc.h +# End Source File +# Begin Source File + +SOURCE=..\..\7zArcIn.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zBuf.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zBuf.h +# End Source File +# Begin Source File + +SOURCE=..\..\7zCrc.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zCrc.h +# End Source File +# Begin Source File + +SOURCE=..\..\7zCrcOpt.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zDec.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zFile.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zFile.h +# End Source File +# Begin Source File + +SOURCE=..\..\7zStream.c +# End Source File +# Begin Source File + +SOURCE=..\..\7zTypes.h +# End Source File +# Begin Source File + +SOURCE=..\..\Bcj2.c +# End Source File +# Begin Source File + +SOURCE=..\..\Bcj2.h +# End Source File +# Begin Source File + +SOURCE=..\..\Bra.c +# End Source File +# Begin Source File + +SOURCE=..\..\Bra.h +# End Source File +# Begin Source File + +SOURCE=..\..\Bra86.c +# End Source File +# Begin Source File + +SOURCE=..\..\BraIA64.c +# End Source File +# Begin Source File + +SOURCE=..\..\CpuArch.c +# End Source File +# Begin Source File + +SOURCE=..\..\CpuArch.h +# End Source File +# Begin Source File + +SOURCE=..\..\Delta.c +# End Source File +# Begin Source File + +SOURCE=..\..\Delta.h +# End Source File +# Begin Source File + +SOURCE=..\..\DllSecur.c +# End Source File +# Begin Source File + +SOURCE=..\..\DllSecur.h +# End Source File +# Begin Source File + +SOURCE=..\..\Lzma2Dec.c +# End Source File +# Begin Source File + +SOURCE=..\..\Lzma2Dec.h +# End Source File +# Begin Source File + +SOURCE=..\..\LzmaDec.c +# End Source File +# Begin Source File + +SOURCE=..\..\LzmaDec.h +# End Source File +# End Group +# Begin Group "Spec" + +# PROP Default_Filter "" +# Begin Source File + +SOURCE=.\Precomp.c +# ADD CPP /Yc"Precomp.h" +# End Source File +# Begin Source File + +SOURCE=.\Precomp.h +# End Source File +# End Group +# Begin Source File + +SOURCE=.\SfxSetup.c +# End Source File +# End Target +# End Project diff --git a/deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.dsw b/deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.dsw index 128fcdd3e..ea2311129 100644 --- a/deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.dsw +++ b/deps/LZMA-SDK/C/Util/SfxSetup/SfxSetup.dsw @@ -1,29 +1,29 @@ -Microsoft Developer Studio Workspace File, Format Version 6.00 -# WARNING: DO NOT EDIT OR DELETE THIS WORKSPACE FILE! - -############################################################################### - -Project: "SfxSetup"=.\SfxSetup.dsp - Package Owner=<4> - -Package=<5> -{{{ -}}} - -Package=<4> -{{{ -}}} - -############################################################################### - -Global: - -Package=<5> -{{{ -}}} - -Package=<3> -{{{ -}}} - -############################################################################### - +Microsoft Developer Studio Workspace File, Format Version 6.00 +# WARNING: DO NOT EDIT OR DELETE THIS WORKSPACE FILE! + +############################################################################### + +Project: "SfxSetup"=.\SfxSetup.dsp - Package Owner=<4> + +Package=<5> +{{{ +}}} + +Package=<4> +{{{ +}}} + +############################################################################### + +Global: + +Package=<5> +{{{ +}}} + +Package=<3> +{{{ +}}} + +############################################################################### + diff --git a/deps/LZMA-SDK/C/Util/SfxSetup/makefile b/deps/LZMA-SDK/C/Util/SfxSetup/makefile index c9f59ccd3..544da67df 100644 --- a/deps/LZMA-SDK/C/Util/SfxSetup/makefile +++ b/deps/LZMA-SDK/C/Util/SfxSetup/makefile @@ -1,37 +1,37 @@ -PROG = 7zS2.sfx -MY_FIXED = 1 - -C_OBJS = \ - $O\7zAlloc.obj \ - $O\7zArcIn.obj \ - $O\7zBuf.obj \ - $O\7zBuf2.obj \ - $O\7zCrc.obj \ - $O\7zCrcOpt.obj \ - $O\7zFile.obj \ - $O\7zDec.obj \ - $O\7zStream.obj \ - $O\Bcj2.obj \ - $O\Bra.obj \ - $O\Bra86.obj \ - $O\BraIA64.obj \ - $O\CpuArch.obj \ - $O\Delta.obj \ - $O\DllSecur.obj \ - $O\Lzma2Dec.obj \ - $O\LzmaDec.obj \ - -7Z_OBJS = \ - $O\SfxSetup.obj \ - -OBJS = \ - $(7Z_OBJS) \ - $(C_OBJS) \ - $O\resource.res - -!include "../../../CPP/Build.mak" - -$(7Z_OBJS): $(*B).c - $(COMPL_O1) -$(C_OBJS): ../../$(*B).c - $(COMPL_O1) +PROG = 7zS2.sfx +MY_FIXED = 1 + +C_OBJS = \ + $O\7zAlloc.obj \ + $O\7zArcIn.obj \ + $O\7zBuf.obj \ + $O\7zBuf2.obj \ + $O\7zCrc.obj \ + $O\7zCrcOpt.obj \ + $O\7zFile.obj \ + $O\7zDec.obj \ + $O\7zStream.obj \ + $O\Bcj2.obj \ + $O\Bra.obj \ + $O\Bra86.obj \ + $O\BraIA64.obj \ + $O\CpuArch.obj \ + $O\Delta.obj \ + $O\DllSecur.obj \ + $O\Lzma2Dec.obj \ + $O\LzmaDec.obj \ + +7Z_OBJS = \ + $O\SfxSetup.obj \ + +OBJS = \ + $(7Z_OBJS) \ + $(C_OBJS) \ + $O\resource.res + +!include "../../../CPP/Build.mak" + +$(7Z_OBJS): $(*B).c + $(COMPL_O1) +$(C_OBJS): ../../$(*B).c + $(COMPL_O1) diff --git a/deps/LZMA-SDK/C/Util/SfxSetup/makefile_con b/deps/LZMA-SDK/C/Util/SfxSetup/makefile_con index 6f604ed81..d0f835254 100644 --- a/deps/LZMA-SDK/C/Util/SfxSetup/makefile_con +++ b/deps/LZMA-SDK/C/Util/SfxSetup/makefile_con @@ -1,38 +1,38 @@ -PROG = 7zS2con.sfx -MY_FIXED = 1 -CFLAGS = $(CFLAGS) -D_CONSOLE - -C_OBJS = \ - $O\7zAlloc.obj \ - $O\7zArcIn.obj \ - $O\7zBuf.obj \ - $O\7zBuf2.obj \ - $O\7zCrc.obj \ - $O\7zCrcOpt.obj \ - $O\7zFile.obj \ - $O\7zDec.obj \ - $O\7zStream.obj \ - $O\Bcj2.obj \ - $O\Bra.obj \ - $O\Bra86.obj \ - $O\BraIA64.obj \ - $O\CpuArch.obj \ - $O\Delta.obj \ - $O\DllSecur.obj \ - $O\Lzma2Dec.obj \ - $O\LzmaDec.obj \ - -7Z_OBJS = \ - $O\SfxSetup.obj \ - -OBJS = \ - $(7Z_OBJS) \ - $(C_OBJS) \ - $O\resource.res - -!include "../../../CPP/Build.mak" - -$(7Z_OBJS): $(*B).c - $(COMPL_O1) -$(C_OBJS): ../../$(*B).c - $(COMPL_O1) +PROG = 7zS2con.sfx +MY_FIXED = 1 +CFLAGS = $(CFLAGS) -D_CONSOLE + +C_OBJS = \ + $O\7zAlloc.obj \ + $O\7zArcIn.obj \ + $O\7zBuf.obj \ + $O\7zBuf2.obj \ + $O\7zCrc.obj \ + $O\7zCrcOpt.obj \ + $O\7zFile.obj \ + $O\7zDec.obj \ + $O\7zStream.obj \ + $O\Bcj2.obj \ + $O\Bra.obj \ + $O\Bra86.obj \ + $O\BraIA64.obj \ + $O\CpuArch.obj \ + $O\Delta.obj \ + $O\DllSecur.obj \ + $O\Lzma2Dec.obj \ + $O\LzmaDec.obj \ + +7Z_OBJS = \ + $O\SfxSetup.obj \ + +OBJS = \ + $(7Z_OBJS) \ + $(C_OBJS) \ + $O\resource.res + +!include "../../../CPP/Build.mak" + +$(7Z_OBJS): $(*B).c + $(COMPL_O1) +$(C_OBJS): ../../$(*B).c + $(COMPL_O1) diff --git a/deps/LZMA-SDK/C/Util/SfxSetup/resource.rc b/deps/LZMA-SDK/C/Util/SfxSetup/resource.rc index 64f4e2ce7..0c1637f23 100644 --- a/deps/LZMA-SDK/C/Util/SfxSetup/resource.rc +++ b/deps/LZMA-SDK/C/Util/SfxSetup/resource.rc @@ -1,5 +1,5 @@ -#include "../../7zVersion.rc" - -MY_VERSION_INFO_APP("7z Setup SFX small", "7zS2.sfx") - -1 ICON "setup.ico" +#include "../../7zVersion.rc" + +MY_VERSION_INFO_APP("7z Setup SFX small", "7zS2.sfx") + +1 ICON "setup.ico" diff --git a/deps/LZMA-SDK/C/Xz.c b/deps/LZMA-SDK/C/Xz.c index d6e2596a9..7c53b6007 100644 --- a/deps/LZMA-SDK/C/Xz.c +++ b/deps/LZMA-SDK/C/Xz.c @@ -1,90 +1,90 @@ -/* Xz.c - Xz -2021-02-09 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "7zCrc.h" -#include "CpuArch.h" -#include "Xz.h" -#include "XzCrc64.h" - -const Byte XZ_SIG[XZ_SIG_SIZE] = { 0xFD, '7', 'z', 'X', 'Z', 0 }; -/* const Byte XZ_FOOTER_SIG[XZ_FOOTER_SIG_SIZE] = { 'Y', 'Z' }; */ - -unsigned Xz_WriteVarInt(Byte *buf, UInt64 v) -{ - unsigned i = 0; - do - { - buf[i++] = (Byte)((v & 0x7F) | 0x80); - v >>= 7; - } - while (v != 0); - buf[(size_t)i - 1] &= 0x7F; - return i; -} - -void Xz_Construct(CXzStream *p) -{ - p->numBlocks = 0; - p->blocks = NULL; - p->flags = 0; -} - -void Xz_Free(CXzStream *p, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, p->blocks); - p->numBlocks = 0; - p->blocks = NULL; -} - -unsigned XzFlags_GetCheckSize(CXzStreamFlags f) -{ - unsigned t = XzFlags_GetCheckType(f); - return (t == 0) ? 0 : ((unsigned)4 << ((t - 1) / 3)); -} - -void XzCheck_Init(CXzCheck *p, unsigned mode) -{ - p->mode = mode; - switch (mode) - { - case XZ_CHECK_CRC32: p->crc = CRC_INIT_VAL; break; - case XZ_CHECK_CRC64: p->crc64 = CRC64_INIT_VAL; break; - case XZ_CHECK_SHA256: Sha256_Init(&p->sha); break; - } -} - -void XzCheck_Update(CXzCheck *p, const void *data, size_t size) -{ - switch (p->mode) - { - case XZ_CHECK_CRC32: p->crc = CrcUpdate(p->crc, data, size); break; - case XZ_CHECK_CRC64: p->crc64 = Crc64Update(p->crc64, data, size); break; - case XZ_CHECK_SHA256: Sha256_Update(&p->sha, (const Byte *)data, size); break; - } -} - -int XzCheck_Final(CXzCheck *p, Byte *digest) -{ - switch (p->mode) - { - case XZ_CHECK_CRC32: - SetUi32(digest, CRC_GET_DIGEST(p->crc)); - break; - case XZ_CHECK_CRC64: - { - int i; - UInt64 v = CRC64_GET_DIGEST(p->crc64); - for (i = 0; i < 8; i++, v >>= 8) - digest[i] = (Byte)(v & 0xFF); - break; - } - case XZ_CHECK_SHA256: - Sha256_Final(&p->sha, digest); - break; - default: - return 0; - } - return 1; -} +/* Xz.c - Xz +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "7zCrc.h" +#include "CpuArch.h" +#include "Xz.h" +#include "XzCrc64.h" + +const Byte XZ_SIG[XZ_SIG_SIZE] = { 0xFD, '7', 'z', 'X', 'Z', 0 }; +/* const Byte XZ_FOOTER_SIG[XZ_FOOTER_SIG_SIZE] = { 'Y', 'Z' }; */ + +unsigned Xz_WriteVarInt(Byte *buf, UInt64 v) +{ + unsigned i = 0; + do + { + buf[i++] = (Byte)((v & 0x7F) | 0x80); + v >>= 7; + } + while (v != 0); + buf[(size_t)i - 1] &= 0x7F; + return i; +} + +void Xz_Construct(CXzStream *p) +{ + p->numBlocks = 0; + p->blocks = NULL; + p->flags = 0; +} + +void Xz_Free(CXzStream *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->blocks); + p->numBlocks = 0; + p->blocks = NULL; +} + +unsigned XzFlags_GetCheckSize(CXzStreamFlags f) +{ + unsigned t = XzFlags_GetCheckType(f); + return (t == 0) ? 0 : ((unsigned)4 << ((t - 1) / 3)); +} + +void XzCheck_Init(CXzCheck *p, unsigned mode) +{ + p->mode = mode; + switch (mode) + { + case XZ_CHECK_CRC32: p->crc = CRC_INIT_VAL; break; + case XZ_CHECK_CRC64: p->crc64 = CRC64_INIT_VAL; break; + case XZ_CHECK_SHA256: Sha256_Init(&p->sha); break; + } +} + +void XzCheck_Update(CXzCheck *p, const void *data, size_t size) +{ + switch (p->mode) + { + case XZ_CHECK_CRC32: p->crc = CrcUpdate(p->crc, data, size); break; + case XZ_CHECK_CRC64: p->crc64 = Crc64Update(p->crc64, data, size); break; + case XZ_CHECK_SHA256: Sha256_Update(&p->sha, (const Byte *)data, size); break; + } +} + +int XzCheck_Final(CXzCheck *p, Byte *digest) +{ + switch (p->mode) + { + case XZ_CHECK_CRC32: + SetUi32(digest, CRC_GET_DIGEST(p->crc)); + break; + case XZ_CHECK_CRC64: + { + int i; + UInt64 v = CRC64_GET_DIGEST(p->crc64); + for (i = 0; i < 8; i++, v >>= 8) + digest[i] = (Byte)(v & 0xFF); + break; + } + case XZ_CHECK_SHA256: + Sha256_Final(&p->sha, digest); + break; + default: + return 0; + } + return 1; +} diff --git a/deps/LZMA-SDK/C/Xz.h b/deps/LZMA-SDK/C/Xz.h index cf9458e39..849b944bf 100644 --- a/deps/LZMA-SDK/C/Xz.h +++ b/deps/LZMA-SDK/C/Xz.h @@ -1,517 +1,517 @@ -/* Xz.h - Xz interface -2021-04-01 : Igor Pavlov : Public domain */ - -#ifndef __XZ_H -#define __XZ_H - -#include "Sha256.h" - -EXTERN_C_BEGIN - -#define XZ_ID_Subblock 1 -#define XZ_ID_Delta 3 -#define XZ_ID_X86 4 -#define XZ_ID_PPC 5 -#define XZ_ID_IA64 6 -#define XZ_ID_ARM 7 -#define XZ_ID_ARMT 8 -#define XZ_ID_SPARC 9 -#define XZ_ID_LZMA2 0x21 - -unsigned Xz_ReadVarInt(const Byte *p, size_t maxSize, UInt64 *value); -unsigned Xz_WriteVarInt(Byte *buf, UInt64 v); - -/* ---------- xz block ---------- */ - -#define XZ_BLOCK_HEADER_SIZE_MAX 1024 - -#define XZ_NUM_FILTERS_MAX 4 -#define XZ_BF_NUM_FILTERS_MASK 3 -#define XZ_BF_PACK_SIZE (1 << 6) -#define XZ_BF_UNPACK_SIZE (1 << 7) - -#define XZ_FILTER_PROPS_SIZE_MAX 20 - -typedef struct -{ - UInt64 id; - UInt32 propsSize; - Byte props[XZ_FILTER_PROPS_SIZE_MAX]; -} CXzFilter; - -typedef struct -{ - UInt64 packSize; - UInt64 unpackSize; - Byte flags; - CXzFilter filters[XZ_NUM_FILTERS_MAX]; -} CXzBlock; - -#define XzBlock_GetNumFilters(p) (((unsigned)(p)->flags & XZ_BF_NUM_FILTERS_MASK) + 1) -#define XzBlock_HasPackSize(p) (((p)->flags & XZ_BF_PACK_SIZE) != 0) -#define XzBlock_HasUnpackSize(p) (((p)->flags & XZ_BF_UNPACK_SIZE) != 0) -#define XzBlock_HasUnsupportedFlags(p) (((p)->flags & ~(XZ_BF_NUM_FILTERS_MASK | XZ_BF_PACK_SIZE | XZ_BF_UNPACK_SIZE)) != 0) - -SRes XzBlock_Parse(CXzBlock *p, const Byte *header); -SRes XzBlock_ReadHeader(CXzBlock *p, ISeqInStream *inStream, BoolInt *isIndex, UInt32 *headerSizeRes); - -/* ---------- xz stream ---------- */ - -#define XZ_SIG_SIZE 6 -#define XZ_FOOTER_SIG_SIZE 2 - -extern const Byte XZ_SIG[XZ_SIG_SIZE]; - -/* -extern const Byte XZ_FOOTER_SIG[XZ_FOOTER_SIG_SIZE]; -*/ - -#define XZ_FOOTER_SIG_0 'Y' -#define XZ_FOOTER_SIG_1 'Z' - -#define XZ_STREAM_FLAGS_SIZE 2 -#define XZ_STREAM_CRC_SIZE 4 - -#define XZ_STREAM_HEADER_SIZE (XZ_SIG_SIZE + XZ_STREAM_FLAGS_SIZE + XZ_STREAM_CRC_SIZE) -#define XZ_STREAM_FOOTER_SIZE (XZ_FOOTER_SIG_SIZE + XZ_STREAM_FLAGS_SIZE + XZ_STREAM_CRC_SIZE + 4) - -#define XZ_CHECK_MASK 0xF -#define XZ_CHECK_NO 0 -#define XZ_CHECK_CRC32 1 -#define XZ_CHECK_CRC64 4 -#define XZ_CHECK_SHA256 10 - -typedef struct -{ - unsigned mode; - UInt32 crc; - UInt64 crc64; - CSha256 sha; -} CXzCheck; - -void XzCheck_Init(CXzCheck *p, unsigned mode); -void XzCheck_Update(CXzCheck *p, const void *data, size_t size); -int XzCheck_Final(CXzCheck *p, Byte *digest); - -typedef UInt16 CXzStreamFlags; - -#define XzFlags_IsSupported(f) ((f) <= XZ_CHECK_MASK) -#define XzFlags_GetCheckType(f) ((f) & XZ_CHECK_MASK) -#define XzFlags_HasDataCrc32(f) (Xz_GetCheckType(f) == XZ_CHECK_CRC32) -unsigned XzFlags_GetCheckSize(CXzStreamFlags f); - -SRes Xz_ParseHeader(CXzStreamFlags *p, const Byte *buf); -SRes Xz_ReadHeader(CXzStreamFlags *p, ISeqInStream *inStream); - -typedef struct -{ - UInt64 unpackSize; - UInt64 totalSize; -} CXzBlockSizes; - -typedef struct -{ - CXzStreamFlags flags; - size_t numBlocks; - CXzBlockSizes *blocks; - UInt64 startOffset; -} CXzStream; - -void Xz_Construct(CXzStream *p); -void Xz_Free(CXzStream *p, ISzAllocPtr alloc); - -#define XZ_SIZE_OVERFLOW ((UInt64)(Int64)-1) - -UInt64 Xz_GetUnpackSize(const CXzStream *p); -UInt64 Xz_GetPackSize(const CXzStream *p); - -typedef struct -{ - size_t num; - size_t numAllocated; - CXzStream *streams; -} CXzs; - -void Xzs_Construct(CXzs *p); -void Xzs_Free(CXzs *p, ISzAllocPtr alloc); -SRes Xzs_ReadBackward(CXzs *p, ILookInStream *inStream, Int64 *startOffset, ICompressProgress *progress, ISzAllocPtr alloc); - -UInt64 Xzs_GetNumBlocks(const CXzs *p); -UInt64 Xzs_GetUnpackSize(const CXzs *p); - - -// ECoderStatus values are identical to ELzmaStatus values of LZMA2 decoder - -typedef enum -{ - CODER_STATUS_NOT_SPECIFIED, /* use main error code instead */ - CODER_STATUS_FINISHED_WITH_MARK, /* stream was finished with end mark. */ - CODER_STATUS_NOT_FINISHED, /* stream was not finished */ - CODER_STATUS_NEEDS_MORE_INPUT /* you must provide more input bytes */ -} ECoderStatus; - - -// ECoderFinishMode values are identical to ELzmaFinishMode - -typedef enum -{ - CODER_FINISH_ANY, /* finish at any point */ - CODER_FINISH_END /* block must be finished at the end */ -} ECoderFinishMode; - - -typedef struct _IStateCoder -{ - void *p; - void (*Free)(void *p, ISzAllocPtr alloc); - SRes (*SetProps)(void *p, const Byte *props, size_t propSize, ISzAllocPtr alloc); - void (*Init)(void *p); - SRes (*Code2)(void *p, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, - int srcWasFinished, ECoderFinishMode finishMode, - // int *wasFinished, - ECoderStatus *status); - SizeT (*Filter)(void *p, Byte *data, SizeT size); -} IStateCoder; - - - -#define MIXCODER_NUM_FILTERS_MAX 4 - -typedef struct -{ - ISzAllocPtr alloc; - Byte *buf; - unsigned numCoders; - - Byte *outBuf; - size_t outBufSize; - size_t outWritten; // is equal to lzmaDecoder.dicPos (in outBuf mode) - BoolInt wasFinished; - SRes res; - ECoderStatus status; - // BoolInt SingleBufMode; - - int finished[MIXCODER_NUM_FILTERS_MAX - 1]; - size_t pos[MIXCODER_NUM_FILTERS_MAX - 1]; - size_t size[MIXCODER_NUM_FILTERS_MAX - 1]; - UInt64 ids[MIXCODER_NUM_FILTERS_MAX]; - SRes results[MIXCODER_NUM_FILTERS_MAX]; - IStateCoder coders[MIXCODER_NUM_FILTERS_MAX]; -} CMixCoder; - - -typedef enum -{ - XZ_STATE_STREAM_HEADER, - XZ_STATE_STREAM_INDEX, - XZ_STATE_STREAM_INDEX_CRC, - XZ_STATE_STREAM_FOOTER, - XZ_STATE_STREAM_PADDING, - XZ_STATE_BLOCK_HEADER, - XZ_STATE_BLOCK, - XZ_STATE_BLOCK_FOOTER -} EXzState; - - -typedef struct -{ - EXzState state; - UInt32 pos; - unsigned alignPos; - unsigned indexPreSize; - - CXzStreamFlags streamFlags; - - UInt32 blockHeaderSize; - UInt64 packSize; - UInt64 unpackSize; - - UInt64 numBlocks; // number of finished blocks in current stream - UInt64 indexSize; - UInt64 indexPos; - UInt64 padSize; - - UInt64 numStartedStreams; - UInt64 numFinishedStreams; - UInt64 numTotalBlocks; - - UInt32 crc; - CMixCoder decoder; - CXzBlock block; - CXzCheck check; - CSha256 sha; - - BoolInt parseMode; - BoolInt headerParsedOk; - BoolInt decodeToStreamSignature; - unsigned decodeOnlyOneBlock; - - Byte *outBuf; - size_t outBufSize; - size_t outDataWritten; // the size of data in (outBuf) that were fully unpacked - - Byte shaDigest[SHA256_DIGEST_SIZE]; - Byte buf[XZ_BLOCK_HEADER_SIZE_MAX]; -} CXzUnpacker; - -/* alloc : aligned for cache line allocation is better */ -void XzUnpacker_Construct(CXzUnpacker *p, ISzAllocPtr alloc); -void XzUnpacker_Init(CXzUnpacker *p); -void XzUnpacker_SetOutBuf(CXzUnpacker *p, Byte *outBuf, size_t outBufSize); -void XzUnpacker_Free(CXzUnpacker *p); - -/* - XzUnpacker - The sequence for decoding functions: - { - XzUnpacker_Construct() - [Decoding_Calls] - XzUnpacker_Free() - } - - [Decoding_Calls] - - There are 3 types of interfaces for [Decoding_Calls] calls: - - Interface-1 : Partial output buffers: - { - XzUnpacker_Init() - for() - { - XzUnpacker_Code(); - } - XzUnpacker_IsStreamWasFinished() - } - - Interface-2 : Direct output buffer: - Use it, if you know exact size of decoded data, and you need - whole xz unpacked data in one output buffer. - xz unpacker doesn't allocate additional buffer for lzma2 dictionary in that mode. - { - XzUnpacker_Init() - XzUnpacker_SetOutBufMode(); // to set output buffer and size - for() - { - XzUnpacker_Code(); // (dest = NULL) in XzUnpacker_Code() - } - XzUnpacker_IsStreamWasFinished() - } - - Interface-3 : Direct output buffer : One call full decoding - It unpacks whole input buffer to output buffer in one call. - It uses Interface-2 internally. - { - XzUnpacker_CodeFull() - XzUnpacker_IsStreamWasFinished() - } -*/ - -/* -finishMode: - It has meaning only if the decoding reaches output limit (*destLen). - CODER_FINISH_ANY - use smallest number of input bytes - CODER_FINISH_END - read EndOfStream marker after decoding - -Returns: - SZ_OK - status: - CODER_STATUS_NOT_FINISHED, - CODER_STATUS_NEEDS_MORE_INPUT - the decoder can return it in two cases: - 1) it needs more input data to finish current xz stream - 2) xz stream was finished successfully. But the decoder supports multiple - concatented xz streams. So it expects more input data for new xz streams. - Call XzUnpacker_IsStreamWasFinished() to check that latest xz stream was finished successfully. - - SZ_ERROR_MEM - Memory allocation error - SZ_ERROR_DATA - Data error - SZ_ERROR_UNSUPPORTED - Unsupported method or method properties - SZ_ERROR_CRC - CRC error - // SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src). - - SZ_ERROR_NO_ARCHIVE - the error with xz Stream Header with one of the following reasons: - - xz Stream Signature failure - - CRC32 of xz Stream Header is failed - - The size of Stream padding is not multiple of four bytes. - It's possible to get that error, if xz stream was finished and the stream - contains some another data. In that case you can call XzUnpacker_GetExtraSize() - function to get real size of xz stream. -*/ - - -SRes XzUnpacker_Code(CXzUnpacker *p, Byte *dest, SizeT *destLen, - const Byte *src, SizeT *srcLen, int srcFinished, - ECoderFinishMode finishMode, ECoderStatus *status); - -SRes XzUnpacker_CodeFull(CXzUnpacker *p, Byte *dest, SizeT *destLen, - const Byte *src, SizeT *srcLen, - ECoderFinishMode finishMode, ECoderStatus *status); - -/* -If you decode full xz stream(s), then you can call XzUnpacker_IsStreamWasFinished() -after successful XzUnpacker_CodeFull() or after last call of XzUnpacker_Code(). -*/ - -BoolInt XzUnpacker_IsStreamWasFinished(const CXzUnpacker *p); - -/* -XzUnpacker_GetExtraSize() returns then number of unconfirmed bytes, - if it's in (XZ_STATE_STREAM_HEADER) state or in (XZ_STATE_STREAM_PADDING) state. -These bytes can be some data after xz archive, or -it can be start of new xz stream. - -Call XzUnpacker_GetExtraSize() after XzUnpacker_Code() function to detect real size of -xz stream in two cases, if XzUnpacker_Code() returns: - res == SZ_OK && status == CODER_STATUS_NEEDS_MORE_INPUT - res == SZ_ERROR_NO_ARCHIVE -*/ - -UInt64 XzUnpacker_GetExtraSize(const CXzUnpacker *p); - - -/* - for random block decoding: - XzUnpacker_Init(); - set CXzUnpacker::streamFlags - XzUnpacker_PrepareToRandomBlockDecoding() - loop - { - XzUnpacker_Code() - XzUnpacker_IsBlockFinished() - } -*/ - -void XzUnpacker_PrepareToRandomBlockDecoding(CXzUnpacker *p); -BoolInt XzUnpacker_IsBlockFinished(const CXzUnpacker *p); - -#define XzUnpacker_GetPackSizeForIndex(p) ((p)->packSize + (p)->blockHeaderSize + XzFlags_GetCheckSize((p)->streamFlags)) - - - - - - -/* ---- Single-Thread and Multi-Thread xz Decoding with Input/Output Streams ---- */ - -/* - if (CXzDecMtProps::numThreads > 1), the decoder can try to use - Multi-Threading. The decoder analyses xz block header, and if - there are pack size and unpack size values stored in xz block header, - the decoder reads compressed data of block to internal buffers, - and then it can start parallel decoding, if there are another blocks. - The decoder can switch back to Single-Thread decoding after some conditions. - - The sequence of calls for xz decoding with in/out Streams: - { - XzDecMt_Create() - XzDecMtProps_Init(XzDecMtProps) to set default values of properties - // then you can change some XzDecMtProps parameters with required values - // here you can set the number of threads and (memUseMax) - the maximum - Memory usage for multithreading decoding. - for() - { - XzDecMt_Decode() // one call per one file - } - XzDecMt_Destroy() - } -*/ - - -typedef struct -{ - size_t inBufSize_ST; // size of input buffer for Single-Thread decoding - size_t outStep_ST; // size of output buffer for Single-Thread decoding - BoolInt ignoreErrors; // if set to 1, the decoder can ignore some errors and it skips broken parts of data. - - #ifndef _7ZIP_ST - unsigned numThreads; // the number of threads for Multi-Thread decoding. if (umThreads == 1) it will use Single-thread decoding - size_t inBufSize_MT; // size of small input data buffers for Multi-Thread decoding. Big number of such small buffers can be created - size_t memUseMax; // the limit of total memory usage for Multi-Thread decoding. - // it's recommended to set (memUseMax) manually to value that is smaller of total size of RAM in computer. - #endif -} CXzDecMtProps; - -void XzDecMtProps_Init(CXzDecMtProps *p); - - -typedef void * CXzDecMtHandle; - -/* - alloc : XzDecMt uses CAlignOffsetAlloc internally for addresses allocated by (alloc). - allocMid : for big allocations, aligned allocation is better -*/ - -CXzDecMtHandle XzDecMt_Create(ISzAllocPtr alloc, ISzAllocPtr allocMid); -void XzDecMt_Destroy(CXzDecMtHandle p); - - -typedef struct -{ - Byte UnpackSize_Defined; - Byte NumStreams_Defined; - Byte NumBlocks_Defined; - - Byte DataAfterEnd; // there are some additional data after good xz streams, and that data is not new xz stream. - Byte DecodingTruncated; // Decoding was Truncated, we need only partial output data - - UInt64 InSize; // pack size processed. That value doesn't include the data after - // end of xz stream, if that data was not correct - UInt64 OutSize; - - UInt64 NumStreams; - UInt64 NumBlocks; - - SRes DecodeRes; // the error code of xz streams data decoding - SRes ReadRes; // error code from ISeqInStream:Read() - SRes ProgressRes; // error code from ICompressProgress:Progress() - - SRes CombinedRes; // Combined result error code that shows main rusult - // = S_OK, if there is no error. - // but check also (DataAfterEnd) that can show additional minor errors. - - SRes CombinedRes_Type; // = SZ_ERROR_READ, if error from ISeqInStream - // = SZ_ERROR_PROGRESS, if error from ICompressProgress - // = SZ_ERROR_WRITE, if error from ISeqOutStream - // = SZ_ERROR_* codes for decoding -} CXzStatInfo; - -void XzStatInfo_Clear(CXzStatInfo *p); - -/* - -XzDecMt_Decode() -SRes: it's combined decoding result. It also is equal to stat->CombinedRes. - - SZ_OK - no error - check also output value in (stat->DataAfterEnd) - that can show additional possible error - - SZ_ERROR_MEM - Memory allocation error - SZ_ERROR_NO_ARCHIVE - is not xz archive - SZ_ERROR_ARCHIVE - Headers error - SZ_ERROR_DATA - Data Error - SZ_ERROR_UNSUPPORTED - Unsupported method or method properties - SZ_ERROR_CRC - CRC Error - SZ_ERROR_INPUT_EOF - it needs more input data - SZ_ERROR_WRITE - ISeqOutStream error - (SZ_ERROR_READ) - ISeqInStream errors - (SZ_ERROR_PROGRESS) - ICompressProgress errors - // SZ_ERROR_THREAD - error in multi-threading functions - MY_SRes_HRESULT_FROM_WRes(WRes_error) - error in multi-threading function -*/ - -SRes XzDecMt_Decode(CXzDecMtHandle p, - const CXzDecMtProps *props, - const UInt64 *outDataSize, // NULL means undefined - int finishMode, // 0 - partial unpacking is allowed, 1 - xz stream(s) must be finished - ISeqOutStream *outStream, - // Byte *outBuf, size_t *outBufSize, - ISeqInStream *inStream, - // const Byte *inData, size_t inDataSize, - CXzStatInfo *stat, // out: decoding results and statistics - int *isMT, // out: 0 means that ST (Single-Thread) version was used - // 1 means that MT (Multi-Thread) version was used - ICompressProgress *progress); - -EXTERN_C_END - -#endif +/* Xz.h - Xz interface +2021-04-01 : Igor Pavlov : Public domain */ + +#ifndef __XZ_H +#define __XZ_H + +#include "Sha256.h" + +EXTERN_C_BEGIN + +#define XZ_ID_Subblock 1 +#define XZ_ID_Delta 3 +#define XZ_ID_X86 4 +#define XZ_ID_PPC 5 +#define XZ_ID_IA64 6 +#define XZ_ID_ARM 7 +#define XZ_ID_ARMT 8 +#define XZ_ID_SPARC 9 +#define XZ_ID_LZMA2 0x21 + +unsigned Xz_ReadVarInt(const Byte *p, size_t maxSize, UInt64 *value); +unsigned Xz_WriteVarInt(Byte *buf, UInt64 v); + +/* ---------- xz block ---------- */ + +#define XZ_BLOCK_HEADER_SIZE_MAX 1024 + +#define XZ_NUM_FILTERS_MAX 4 +#define XZ_BF_NUM_FILTERS_MASK 3 +#define XZ_BF_PACK_SIZE (1 << 6) +#define XZ_BF_UNPACK_SIZE (1 << 7) + +#define XZ_FILTER_PROPS_SIZE_MAX 20 + +typedef struct +{ + UInt64 id; + UInt32 propsSize; + Byte props[XZ_FILTER_PROPS_SIZE_MAX]; +} CXzFilter; + +typedef struct +{ + UInt64 packSize; + UInt64 unpackSize; + Byte flags; + CXzFilter filters[XZ_NUM_FILTERS_MAX]; +} CXzBlock; + +#define XzBlock_GetNumFilters(p) (((unsigned)(p)->flags & XZ_BF_NUM_FILTERS_MASK) + 1) +#define XzBlock_HasPackSize(p) (((p)->flags & XZ_BF_PACK_SIZE) != 0) +#define XzBlock_HasUnpackSize(p) (((p)->flags & XZ_BF_UNPACK_SIZE) != 0) +#define XzBlock_HasUnsupportedFlags(p) (((p)->flags & ~(XZ_BF_NUM_FILTERS_MASK | XZ_BF_PACK_SIZE | XZ_BF_UNPACK_SIZE)) != 0) + +SRes XzBlock_Parse(CXzBlock *p, const Byte *header); +SRes XzBlock_ReadHeader(CXzBlock *p, ISeqInStream *inStream, BoolInt *isIndex, UInt32 *headerSizeRes); + +/* ---------- xz stream ---------- */ + +#define XZ_SIG_SIZE 6 +#define XZ_FOOTER_SIG_SIZE 2 + +extern const Byte XZ_SIG[XZ_SIG_SIZE]; + +/* +extern const Byte XZ_FOOTER_SIG[XZ_FOOTER_SIG_SIZE]; +*/ + +#define XZ_FOOTER_SIG_0 'Y' +#define XZ_FOOTER_SIG_1 'Z' + +#define XZ_STREAM_FLAGS_SIZE 2 +#define XZ_STREAM_CRC_SIZE 4 + +#define XZ_STREAM_HEADER_SIZE (XZ_SIG_SIZE + XZ_STREAM_FLAGS_SIZE + XZ_STREAM_CRC_SIZE) +#define XZ_STREAM_FOOTER_SIZE (XZ_FOOTER_SIG_SIZE + XZ_STREAM_FLAGS_SIZE + XZ_STREAM_CRC_SIZE + 4) + +#define XZ_CHECK_MASK 0xF +#define XZ_CHECK_NO 0 +#define XZ_CHECK_CRC32 1 +#define XZ_CHECK_CRC64 4 +#define XZ_CHECK_SHA256 10 + +typedef struct +{ + unsigned mode; + UInt32 crc; + UInt64 crc64; + CSha256 sha; +} CXzCheck; + +void XzCheck_Init(CXzCheck *p, unsigned mode); +void XzCheck_Update(CXzCheck *p, const void *data, size_t size); +int XzCheck_Final(CXzCheck *p, Byte *digest); + +typedef UInt16 CXzStreamFlags; + +#define XzFlags_IsSupported(f) ((f) <= XZ_CHECK_MASK) +#define XzFlags_GetCheckType(f) ((f) & XZ_CHECK_MASK) +#define XzFlags_HasDataCrc32(f) (Xz_GetCheckType(f) == XZ_CHECK_CRC32) +unsigned XzFlags_GetCheckSize(CXzStreamFlags f); + +SRes Xz_ParseHeader(CXzStreamFlags *p, const Byte *buf); +SRes Xz_ReadHeader(CXzStreamFlags *p, ISeqInStream *inStream); + +typedef struct +{ + UInt64 unpackSize; + UInt64 totalSize; +} CXzBlockSizes; + +typedef struct +{ + CXzStreamFlags flags; + size_t numBlocks; + CXzBlockSizes *blocks; + UInt64 startOffset; +} CXzStream; + +void Xz_Construct(CXzStream *p); +void Xz_Free(CXzStream *p, ISzAllocPtr alloc); + +#define XZ_SIZE_OVERFLOW ((UInt64)(Int64)-1) + +UInt64 Xz_GetUnpackSize(const CXzStream *p); +UInt64 Xz_GetPackSize(const CXzStream *p); + +typedef struct +{ + size_t num; + size_t numAllocated; + CXzStream *streams; +} CXzs; + +void Xzs_Construct(CXzs *p); +void Xzs_Free(CXzs *p, ISzAllocPtr alloc); +SRes Xzs_ReadBackward(CXzs *p, ILookInStream *inStream, Int64 *startOffset, ICompressProgress *progress, ISzAllocPtr alloc); + +UInt64 Xzs_GetNumBlocks(const CXzs *p); +UInt64 Xzs_GetUnpackSize(const CXzs *p); + + +// ECoderStatus values are identical to ELzmaStatus values of LZMA2 decoder + +typedef enum +{ + CODER_STATUS_NOT_SPECIFIED, /* use main error code instead */ + CODER_STATUS_FINISHED_WITH_MARK, /* stream was finished with end mark. */ + CODER_STATUS_NOT_FINISHED, /* stream was not finished */ + CODER_STATUS_NEEDS_MORE_INPUT /* you must provide more input bytes */ +} ECoderStatus; + + +// ECoderFinishMode values are identical to ELzmaFinishMode + +typedef enum +{ + CODER_FINISH_ANY, /* finish at any point */ + CODER_FINISH_END /* block must be finished at the end */ +} ECoderFinishMode; + + +typedef struct _IStateCoder +{ + void *p; + void (*Free)(void *p, ISzAllocPtr alloc); + SRes (*SetProps)(void *p, const Byte *props, size_t propSize, ISzAllocPtr alloc); + void (*Init)(void *p); + SRes (*Code2)(void *p, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + int srcWasFinished, ECoderFinishMode finishMode, + // int *wasFinished, + ECoderStatus *status); + SizeT (*Filter)(void *p, Byte *data, SizeT size); +} IStateCoder; + + + +#define MIXCODER_NUM_FILTERS_MAX 4 + +typedef struct +{ + ISzAllocPtr alloc; + Byte *buf; + unsigned numCoders; + + Byte *outBuf; + size_t outBufSize; + size_t outWritten; // is equal to lzmaDecoder.dicPos (in outBuf mode) + BoolInt wasFinished; + SRes res; + ECoderStatus status; + // BoolInt SingleBufMode; + + int finished[MIXCODER_NUM_FILTERS_MAX - 1]; + size_t pos[MIXCODER_NUM_FILTERS_MAX - 1]; + size_t size[MIXCODER_NUM_FILTERS_MAX - 1]; + UInt64 ids[MIXCODER_NUM_FILTERS_MAX]; + SRes results[MIXCODER_NUM_FILTERS_MAX]; + IStateCoder coders[MIXCODER_NUM_FILTERS_MAX]; +} CMixCoder; + + +typedef enum +{ + XZ_STATE_STREAM_HEADER, + XZ_STATE_STREAM_INDEX, + XZ_STATE_STREAM_INDEX_CRC, + XZ_STATE_STREAM_FOOTER, + XZ_STATE_STREAM_PADDING, + XZ_STATE_BLOCK_HEADER, + XZ_STATE_BLOCK, + XZ_STATE_BLOCK_FOOTER +} EXzState; + + +typedef struct +{ + EXzState state; + UInt32 pos; + unsigned alignPos; + unsigned indexPreSize; + + CXzStreamFlags streamFlags; + + UInt32 blockHeaderSize; + UInt64 packSize; + UInt64 unpackSize; + + UInt64 numBlocks; // number of finished blocks in current stream + UInt64 indexSize; + UInt64 indexPos; + UInt64 padSize; + + UInt64 numStartedStreams; + UInt64 numFinishedStreams; + UInt64 numTotalBlocks; + + UInt32 crc; + CMixCoder decoder; + CXzBlock block; + CXzCheck check; + CSha256 sha; + + BoolInt parseMode; + BoolInt headerParsedOk; + BoolInt decodeToStreamSignature; + unsigned decodeOnlyOneBlock; + + Byte *outBuf; + size_t outBufSize; + size_t outDataWritten; // the size of data in (outBuf) that were fully unpacked + + Byte shaDigest[SHA256_DIGEST_SIZE]; + Byte buf[XZ_BLOCK_HEADER_SIZE_MAX]; +} CXzUnpacker; + +/* alloc : aligned for cache line allocation is better */ +void XzUnpacker_Construct(CXzUnpacker *p, ISzAllocPtr alloc); +void XzUnpacker_Init(CXzUnpacker *p); +void XzUnpacker_SetOutBuf(CXzUnpacker *p, Byte *outBuf, size_t outBufSize); +void XzUnpacker_Free(CXzUnpacker *p); + +/* + XzUnpacker + The sequence for decoding functions: + { + XzUnpacker_Construct() + [Decoding_Calls] + XzUnpacker_Free() + } + + [Decoding_Calls] + + There are 3 types of interfaces for [Decoding_Calls] calls: + + Interface-1 : Partial output buffers: + { + XzUnpacker_Init() + for() + { + XzUnpacker_Code(); + } + XzUnpacker_IsStreamWasFinished() + } + + Interface-2 : Direct output buffer: + Use it, if you know exact size of decoded data, and you need + whole xz unpacked data in one output buffer. + xz unpacker doesn't allocate additional buffer for lzma2 dictionary in that mode. + { + XzUnpacker_Init() + XzUnpacker_SetOutBufMode(); // to set output buffer and size + for() + { + XzUnpacker_Code(); // (dest = NULL) in XzUnpacker_Code() + } + XzUnpacker_IsStreamWasFinished() + } + + Interface-3 : Direct output buffer : One call full decoding + It unpacks whole input buffer to output buffer in one call. + It uses Interface-2 internally. + { + XzUnpacker_CodeFull() + XzUnpacker_IsStreamWasFinished() + } +*/ + +/* +finishMode: + It has meaning only if the decoding reaches output limit (*destLen). + CODER_FINISH_ANY - use smallest number of input bytes + CODER_FINISH_END - read EndOfStream marker after decoding + +Returns: + SZ_OK + status: + CODER_STATUS_NOT_FINISHED, + CODER_STATUS_NEEDS_MORE_INPUT - the decoder can return it in two cases: + 1) it needs more input data to finish current xz stream + 2) xz stream was finished successfully. But the decoder supports multiple + concatented xz streams. So it expects more input data for new xz streams. + Call XzUnpacker_IsStreamWasFinished() to check that latest xz stream was finished successfully. + + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_DATA - Data error + SZ_ERROR_UNSUPPORTED - Unsupported method or method properties + SZ_ERROR_CRC - CRC error + // SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src). + + SZ_ERROR_NO_ARCHIVE - the error with xz Stream Header with one of the following reasons: + - xz Stream Signature failure + - CRC32 of xz Stream Header is failed + - The size of Stream padding is not multiple of four bytes. + It's possible to get that error, if xz stream was finished and the stream + contains some another data. In that case you can call XzUnpacker_GetExtraSize() + function to get real size of xz stream. +*/ + + +SRes XzUnpacker_Code(CXzUnpacker *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, int srcFinished, + ECoderFinishMode finishMode, ECoderStatus *status); + +SRes XzUnpacker_CodeFull(CXzUnpacker *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, + ECoderFinishMode finishMode, ECoderStatus *status); + +/* +If you decode full xz stream(s), then you can call XzUnpacker_IsStreamWasFinished() +after successful XzUnpacker_CodeFull() or after last call of XzUnpacker_Code(). +*/ + +BoolInt XzUnpacker_IsStreamWasFinished(const CXzUnpacker *p); + +/* +XzUnpacker_GetExtraSize() returns then number of unconfirmed bytes, + if it's in (XZ_STATE_STREAM_HEADER) state or in (XZ_STATE_STREAM_PADDING) state. +These bytes can be some data after xz archive, or +it can be start of new xz stream. + +Call XzUnpacker_GetExtraSize() after XzUnpacker_Code() function to detect real size of +xz stream in two cases, if XzUnpacker_Code() returns: + res == SZ_OK && status == CODER_STATUS_NEEDS_MORE_INPUT + res == SZ_ERROR_NO_ARCHIVE +*/ + +UInt64 XzUnpacker_GetExtraSize(const CXzUnpacker *p); + + +/* + for random block decoding: + XzUnpacker_Init(); + set CXzUnpacker::streamFlags + XzUnpacker_PrepareToRandomBlockDecoding() + loop + { + XzUnpacker_Code() + XzUnpacker_IsBlockFinished() + } +*/ + +void XzUnpacker_PrepareToRandomBlockDecoding(CXzUnpacker *p); +BoolInt XzUnpacker_IsBlockFinished(const CXzUnpacker *p); + +#define XzUnpacker_GetPackSizeForIndex(p) ((p)->packSize + (p)->blockHeaderSize + XzFlags_GetCheckSize((p)->streamFlags)) + + + + + + +/* ---- Single-Thread and Multi-Thread xz Decoding with Input/Output Streams ---- */ + +/* + if (CXzDecMtProps::numThreads > 1), the decoder can try to use + Multi-Threading. The decoder analyses xz block header, and if + there are pack size and unpack size values stored in xz block header, + the decoder reads compressed data of block to internal buffers, + and then it can start parallel decoding, if there are another blocks. + The decoder can switch back to Single-Thread decoding after some conditions. + + The sequence of calls for xz decoding with in/out Streams: + { + XzDecMt_Create() + XzDecMtProps_Init(XzDecMtProps) to set default values of properties + // then you can change some XzDecMtProps parameters with required values + // here you can set the number of threads and (memUseMax) - the maximum + Memory usage for multithreading decoding. + for() + { + XzDecMt_Decode() // one call per one file + } + XzDecMt_Destroy() + } +*/ + + +typedef struct +{ + size_t inBufSize_ST; // size of input buffer for Single-Thread decoding + size_t outStep_ST; // size of output buffer for Single-Thread decoding + BoolInt ignoreErrors; // if set to 1, the decoder can ignore some errors and it skips broken parts of data. + + #ifndef _7ZIP_ST + unsigned numThreads; // the number of threads for Multi-Thread decoding. if (umThreads == 1) it will use Single-thread decoding + size_t inBufSize_MT; // size of small input data buffers for Multi-Thread decoding. Big number of such small buffers can be created + size_t memUseMax; // the limit of total memory usage for Multi-Thread decoding. + // it's recommended to set (memUseMax) manually to value that is smaller of total size of RAM in computer. + #endif +} CXzDecMtProps; + +void XzDecMtProps_Init(CXzDecMtProps *p); + + +typedef void * CXzDecMtHandle; + +/* + alloc : XzDecMt uses CAlignOffsetAlloc internally for addresses allocated by (alloc). + allocMid : for big allocations, aligned allocation is better +*/ + +CXzDecMtHandle XzDecMt_Create(ISzAllocPtr alloc, ISzAllocPtr allocMid); +void XzDecMt_Destroy(CXzDecMtHandle p); + + +typedef struct +{ + Byte UnpackSize_Defined; + Byte NumStreams_Defined; + Byte NumBlocks_Defined; + + Byte DataAfterEnd; // there are some additional data after good xz streams, and that data is not new xz stream. + Byte DecodingTruncated; // Decoding was Truncated, we need only partial output data + + UInt64 InSize; // pack size processed. That value doesn't include the data after + // end of xz stream, if that data was not correct + UInt64 OutSize; + + UInt64 NumStreams; + UInt64 NumBlocks; + + SRes DecodeRes; // the error code of xz streams data decoding + SRes ReadRes; // error code from ISeqInStream:Read() + SRes ProgressRes; // error code from ICompressProgress:Progress() + + SRes CombinedRes; // Combined result error code that shows main rusult + // = S_OK, if there is no error. + // but check also (DataAfterEnd) that can show additional minor errors. + + SRes CombinedRes_Type; // = SZ_ERROR_READ, if error from ISeqInStream + // = SZ_ERROR_PROGRESS, if error from ICompressProgress + // = SZ_ERROR_WRITE, if error from ISeqOutStream + // = SZ_ERROR_* codes for decoding +} CXzStatInfo; + +void XzStatInfo_Clear(CXzStatInfo *p); + +/* + +XzDecMt_Decode() +SRes: it's combined decoding result. It also is equal to stat->CombinedRes. + + SZ_OK - no error + check also output value in (stat->DataAfterEnd) + that can show additional possible error + + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_NO_ARCHIVE - is not xz archive + SZ_ERROR_ARCHIVE - Headers error + SZ_ERROR_DATA - Data Error + SZ_ERROR_UNSUPPORTED - Unsupported method or method properties + SZ_ERROR_CRC - CRC Error + SZ_ERROR_INPUT_EOF - it needs more input data + SZ_ERROR_WRITE - ISeqOutStream error + (SZ_ERROR_READ) - ISeqInStream errors + (SZ_ERROR_PROGRESS) - ICompressProgress errors + // SZ_ERROR_THREAD - error in multi-threading functions + MY_SRes_HRESULT_FROM_WRes(WRes_error) - error in multi-threading function +*/ + +SRes XzDecMt_Decode(CXzDecMtHandle p, + const CXzDecMtProps *props, + const UInt64 *outDataSize, // NULL means undefined + int finishMode, // 0 - partial unpacking is allowed, 1 - xz stream(s) must be finished + ISeqOutStream *outStream, + // Byte *outBuf, size_t *outBufSize, + ISeqInStream *inStream, + // const Byte *inData, size_t inDataSize, + CXzStatInfo *stat, // out: decoding results and statistics + int *isMT, // out: 0 means that ST (Single-Thread) version was used + // 1 means that MT (Multi-Thread) version was used + ICompressProgress *progress); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/XzCrc64.c b/deps/LZMA-SDK/C/XzCrc64.c index e9ca9ec26..b6d02cbeb 100644 --- a/deps/LZMA-SDK/C/XzCrc64.c +++ b/deps/LZMA-SDK/C/XzCrc64.c @@ -1,86 +1,86 @@ -/* XzCrc64.c -- CRC64 calculation -2017-05-23 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "XzCrc64.h" -#include "CpuArch.h" - -#define kCrc64Poly UINT64_CONST(0xC96C5795D7870F42) - -#ifdef MY_CPU_LE - #define CRC64_NUM_TABLES 4 -#else - #define CRC64_NUM_TABLES 5 - #define CRC_UINT64_SWAP(v) \ - ((v >> 56) \ - | ((v >> 40) & ((UInt64)0xFF << 8)) \ - | ((v >> 24) & ((UInt64)0xFF << 16)) \ - | ((v >> 8) & ((UInt64)0xFF << 24)) \ - | ((v << 8) & ((UInt64)0xFF << 32)) \ - | ((v << 24) & ((UInt64)0xFF << 40)) \ - | ((v << 40) & ((UInt64)0xFF << 48)) \ - | ((v << 56))) - - UInt64 MY_FAST_CALL XzCrc64UpdateT1_BeT4(UInt64 v, const void *data, size_t size, const UInt64 *table); -#endif - -#ifndef MY_CPU_BE - UInt64 MY_FAST_CALL XzCrc64UpdateT4(UInt64 v, const void *data, size_t size, const UInt64 *table); -#endif - -typedef UInt64 (MY_FAST_CALL *CRC64_FUNC)(UInt64 v, const void *data, size_t size, const UInt64 *table); - -static CRC64_FUNC g_Crc64Update; -UInt64 g_Crc64Table[256 * CRC64_NUM_TABLES]; - -UInt64 MY_FAST_CALL Crc64Update(UInt64 v, const void *data, size_t size) -{ - return g_Crc64Update(v, data, size, g_Crc64Table); -} - -UInt64 MY_FAST_CALL Crc64Calc(const void *data, size_t size) -{ - return g_Crc64Update(CRC64_INIT_VAL, data, size, g_Crc64Table) ^ CRC64_INIT_VAL; -} - -void MY_FAST_CALL Crc64GenerateTable() -{ - UInt32 i; - for (i = 0; i < 256; i++) - { - UInt64 r = i; - unsigned j; - for (j = 0; j < 8; j++) - r = (r >> 1) ^ (kCrc64Poly & ((UInt64)0 - (r & 1))); - g_Crc64Table[i] = r; - } - for (i = 256; i < 256 * CRC64_NUM_TABLES; i++) - { - UInt64 r = g_Crc64Table[(size_t)i - 256]; - g_Crc64Table[i] = g_Crc64Table[r & 0xFF] ^ (r >> 8); - } - - #ifdef MY_CPU_LE - - g_Crc64Update = XzCrc64UpdateT4; - - #else - { - #ifndef MY_CPU_BE - UInt32 k = 1; - if (*(const Byte *)&k == 1) - g_Crc64Update = XzCrc64UpdateT4; - else - #endif - { - for (i = 256 * CRC64_NUM_TABLES - 1; i >= 256; i--) - { - UInt64 x = g_Crc64Table[(size_t)i - 256]; - g_Crc64Table[i] = CRC_UINT64_SWAP(x); - } - g_Crc64Update = XzCrc64UpdateT1_BeT4; - } - } - #endif -} +/* XzCrc64.c -- CRC64 calculation +2017-05-23 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "XzCrc64.h" +#include "CpuArch.h" + +#define kCrc64Poly UINT64_CONST(0xC96C5795D7870F42) + +#ifdef MY_CPU_LE + #define CRC64_NUM_TABLES 4 +#else + #define CRC64_NUM_TABLES 5 + #define CRC_UINT64_SWAP(v) \ + ((v >> 56) \ + | ((v >> 40) & ((UInt64)0xFF << 8)) \ + | ((v >> 24) & ((UInt64)0xFF << 16)) \ + | ((v >> 8) & ((UInt64)0xFF << 24)) \ + | ((v << 8) & ((UInt64)0xFF << 32)) \ + | ((v << 24) & ((UInt64)0xFF << 40)) \ + | ((v << 40) & ((UInt64)0xFF << 48)) \ + | ((v << 56))) + + UInt64 MY_FAST_CALL XzCrc64UpdateT1_BeT4(UInt64 v, const void *data, size_t size, const UInt64 *table); +#endif + +#ifndef MY_CPU_BE + UInt64 MY_FAST_CALL XzCrc64UpdateT4(UInt64 v, const void *data, size_t size, const UInt64 *table); +#endif + +typedef UInt64 (MY_FAST_CALL *CRC64_FUNC)(UInt64 v, const void *data, size_t size, const UInt64 *table); + +static CRC64_FUNC g_Crc64Update; +UInt64 g_Crc64Table[256 * CRC64_NUM_TABLES]; + +UInt64 MY_FAST_CALL Crc64Update(UInt64 v, const void *data, size_t size) +{ + return g_Crc64Update(v, data, size, g_Crc64Table); +} + +UInt64 MY_FAST_CALL Crc64Calc(const void *data, size_t size) +{ + return g_Crc64Update(CRC64_INIT_VAL, data, size, g_Crc64Table) ^ CRC64_INIT_VAL; +} + +void MY_FAST_CALL Crc64GenerateTable() +{ + UInt32 i; + for (i = 0; i < 256; i++) + { + UInt64 r = i; + unsigned j; + for (j = 0; j < 8; j++) + r = (r >> 1) ^ (kCrc64Poly & ((UInt64)0 - (r & 1))); + g_Crc64Table[i] = r; + } + for (i = 256; i < 256 * CRC64_NUM_TABLES; i++) + { + UInt64 r = g_Crc64Table[(size_t)i - 256]; + g_Crc64Table[i] = g_Crc64Table[r & 0xFF] ^ (r >> 8); + } + + #ifdef MY_CPU_LE + + g_Crc64Update = XzCrc64UpdateT4; + + #else + { + #ifndef MY_CPU_BE + UInt32 k = 1; + if (*(const Byte *)&k == 1) + g_Crc64Update = XzCrc64UpdateT4; + else + #endif + { + for (i = 256 * CRC64_NUM_TABLES - 1; i >= 256; i--) + { + UInt64 x = g_Crc64Table[(size_t)i - 256]; + g_Crc64Table[i] = CRC_UINT64_SWAP(x); + } + g_Crc64Update = XzCrc64UpdateT1_BeT4; + } + } + #endif +} diff --git a/deps/LZMA-SDK/C/XzCrc64.h b/deps/LZMA-SDK/C/XzCrc64.h index 71b10d57e..08dbc330c 100644 --- a/deps/LZMA-SDK/C/XzCrc64.h +++ b/deps/LZMA-SDK/C/XzCrc64.h @@ -1,26 +1,26 @@ -/* XzCrc64.h -- CRC64 calculation -2013-01-18 : Igor Pavlov : Public domain */ - -#ifndef __XZ_CRC64_H -#define __XZ_CRC64_H - -#include - -#include "7zTypes.h" - -EXTERN_C_BEGIN - -extern UInt64 g_Crc64Table[]; - -void MY_FAST_CALL Crc64GenerateTable(void); - -#define CRC64_INIT_VAL UINT64_CONST(0xFFFFFFFFFFFFFFFF) -#define CRC64_GET_DIGEST(crc) ((crc) ^ CRC64_INIT_VAL) -#define CRC64_UPDATE_BYTE(crc, b) (g_Crc64Table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) - -UInt64 MY_FAST_CALL Crc64Update(UInt64 crc, const void *data, size_t size); -UInt64 MY_FAST_CALL Crc64Calc(const void *data, size_t size); - -EXTERN_C_END - -#endif +/* XzCrc64.h -- CRC64 calculation +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __XZ_CRC64_H +#define __XZ_CRC64_H + +#include + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +extern UInt64 g_Crc64Table[]; + +void MY_FAST_CALL Crc64GenerateTable(void); + +#define CRC64_INIT_VAL UINT64_CONST(0xFFFFFFFFFFFFFFFF) +#define CRC64_GET_DIGEST(crc) ((crc) ^ CRC64_INIT_VAL) +#define CRC64_UPDATE_BYTE(crc, b) (g_Crc64Table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) + +UInt64 MY_FAST_CALL Crc64Update(UInt64 crc, const void *data, size_t size); +UInt64 MY_FAST_CALL Crc64Calc(const void *data, size_t size); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/XzCrc64Opt.c b/deps/LZMA-SDK/C/XzCrc64Opt.c index a0637dd22..93a9ffff5 100644 --- a/deps/LZMA-SDK/C/XzCrc64Opt.c +++ b/deps/LZMA-SDK/C/XzCrc64Opt.c @@ -1,71 +1,71 @@ -/* XzCrc64Opt.c -- CRC64 calculation -2021-02-09 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include "CpuArch.h" - -#ifndef MY_CPU_BE - -#define CRC64_UPDATE_BYTE_2(crc, b) (table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) - -UInt64 MY_FAST_CALL XzCrc64UpdateT4(UInt64 v, const void *data, size_t size, const UInt64 *table); -UInt64 MY_FAST_CALL XzCrc64UpdateT4(UInt64 v, const void *data, size_t size, const UInt64 *table) -{ - const Byte *p = (const Byte *)data; - for (; size > 0 && ((unsigned)(ptrdiff_t)p & 3) != 0; size--, p++) - v = CRC64_UPDATE_BYTE_2(v, *p); - for (; size >= 4; size -= 4, p += 4) - { - UInt32 d = (UInt32)v ^ *(const UInt32 *)(const void *)p; - v = (v >> 32) - ^ (table + 0x300)[((d ) & 0xFF)] - ^ (table + 0x200)[((d >> 8) & 0xFF)] - ^ (table + 0x100)[((d >> 16) & 0xFF)] - ^ (table + 0x000)[((d >> 24))]; - } - for (; size > 0; size--, p++) - v = CRC64_UPDATE_BYTE_2(v, *p); - return v; -} - -#endif - - -#ifndef MY_CPU_LE - -#define CRC_UINT64_SWAP(v) \ - ((v >> 56) \ - | ((v >> 40) & ((UInt64)0xFF << 8)) \ - | ((v >> 24) & ((UInt64)0xFF << 16)) \ - | ((v >> 8) & ((UInt64)0xFF << 24)) \ - | ((v << 8) & ((UInt64)0xFF << 32)) \ - | ((v << 24) & ((UInt64)0xFF << 40)) \ - | ((v << 40) & ((UInt64)0xFF << 48)) \ - | ((v << 56))) - -#define CRC64_UPDATE_BYTE_2_BE(crc, b) (table[(Byte)((crc) >> 56) ^ (b)] ^ ((crc) << 8)) - -UInt64 MY_FAST_CALL XzCrc64UpdateT1_BeT4(UInt64 v, const void *data, size_t size, const UInt64 *table); -UInt64 MY_FAST_CALL XzCrc64UpdateT1_BeT4(UInt64 v, const void *data, size_t size, const UInt64 *table) -{ - const Byte *p = (const Byte *)data; - table += 0x100; - v = CRC_UINT64_SWAP(v); - for (; size > 0 && ((unsigned)(ptrdiff_t)p & 3) != 0; size--, p++) - v = CRC64_UPDATE_BYTE_2_BE(v, *p); - for (; size >= 4; size -= 4, p += 4) - { - UInt32 d = (UInt32)(v >> 32) ^ *(const UInt32 *)(const void *)p; - v = (v << 32) - ^ (table + 0x000)[((d ) & 0xFF)] - ^ (table + 0x100)[((d >> 8) & 0xFF)] - ^ (table + 0x200)[((d >> 16) & 0xFF)] - ^ (table + 0x300)[((d >> 24))]; - } - for (; size > 0; size--, p++) - v = CRC64_UPDATE_BYTE_2_BE(v, *p); - return CRC_UINT64_SWAP(v); -} - -#endif +/* XzCrc64Opt.c -- CRC64 calculation +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" + +#ifndef MY_CPU_BE + +#define CRC64_UPDATE_BYTE_2(crc, b) (table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) + +UInt64 MY_FAST_CALL XzCrc64UpdateT4(UInt64 v, const void *data, size_t size, const UInt64 *table); +UInt64 MY_FAST_CALL XzCrc64UpdateT4(UInt64 v, const void *data, size_t size, const UInt64 *table) +{ + const Byte *p = (const Byte *)data; + for (; size > 0 && ((unsigned)(ptrdiff_t)p & 3) != 0; size--, p++) + v = CRC64_UPDATE_BYTE_2(v, *p); + for (; size >= 4; size -= 4, p += 4) + { + UInt32 d = (UInt32)v ^ *(const UInt32 *)(const void *)p; + v = (v >> 32) + ^ (table + 0x300)[((d ) & 0xFF)] + ^ (table + 0x200)[((d >> 8) & 0xFF)] + ^ (table + 0x100)[((d >> 16) & 0xFF)] + ^ (table + 0x000)[((d >> 24))]; + } + for (; size > 0; size--, p++) + v = CRC64_UPDATE_BYTE_2(v, *p); + return v; +} + +#endif + + +#ifndef MY_CPU_LE + +#define CRC_UINT64_SWAP(v) \ + ((v >> 56) \ + | ((v >> 40) & ((UInt64)0xFF << 8)) \ + | ((v >> 24) & ((UInt64)0xFF << 16)) \ + | ((v >> 8) & ((UInt64)0xFF << 24)) \ + | ((v << 8) & ((UInt64)0xFF << 32)) \ + | ((v << 24) & ((UInt64)0xFF << 40)) \ + | ((v << 40) & ((UInt64)0xFF << 48)) \ + | ((v << 56))) + +#define CRC64_UPDATE_BYTE_2_BE(crc, b) (table[(Byte)((crc) >> 56) ^ (b)] ^ ((crc) << 8)) + +UInt64 MY_FAST_CALL XzCrc64UpdateT1_BeT4(UInt64 v, const void *data, size_t size, const UInt64 *table); +UInt64 MY_FAST_CALL XzCrc64UpdateT1_BeT4(UInt64 v, const void *data, size_t size, const UInt64 *table) +{ + const Byte *p = (const Byte *)data; + table += 0x100; + v = CRC_UINT64_SWAP(v); + for (; size > 0 && ((unsigned)(ptrdiff_t)p & 3) != 0; size--, p++) + v = CRC64_UPDATE_BYTE_2_BE(v, *p); + for (; size >= 4; size -= 4, p += 4) + { + UInt32 d = (UInt32)(v >> 32) ^ *(const UInt32 *)(const void *)p; + v = (v << 32) + ^ (table + 0x000)[((d ) & 0xFF)] + ^ (table + 0x100)[((d >> 8) & 0xFF)] + ^ (table + 0x200)[((d >> 16) & 0xFF)] + ^ (table + 0x300)[((d >> 24))]; + } + for (; size > 0; size--, p++) + v = CRC64_UPDATE_BYTE_2_BE(v, *p); + return CRC_UINT64_SWAP(v); +} + +#endif diff --git a/deps/LZMA-SDK/C/XzDec.c b/deps/LZMA-SDK/C/XzDec.c index 49329f16a..3f96a37f9 100644 --- a/deps/LZMA-SDK/C/XzDec.c +++ b/deps/LZMA-SDK/C/XzDec.c @@ -1,2837 +1,2837 @@ -/* XzDec.c -- Xz Decode -2021-09-04 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -// #include - -// #define XZ_DUMP - -/* #define XZ_DUMP */ - -#ifdef XZ_DUMP -#include -#endif - -// #define SHOW_DEBUG_INFO - -#ifdef SHOW_DEBUG_INFO -#include -#endif - -#ifdef SHOW_DEBUG_INFO -#define PRF(x) x -#else -#define PRF(x) -#endif - -#define PRF_STR(s) PRF(printf("\n" s "\n")) -#define PRF_STR_INT(s, d) PRF(printf("\n" s " %d\n", (unsigned)d)) - -#include -#include - -#include "7zCrc.h" -#include "Alloc.h" -#include "Bra.h" -#include "CpuArch.h" -#include "Delta.h" -#include "Lzma2Dec.h" - -// #define USE_SUBBLOCK - -#ifdef USE_SUBBLOCK -#include "Bcj3Dec.c" -#include "SbDec.h" -#endif - -#include "Xz.h" - -#define XZ_CHECK_SIZE_MAX 64 - -#define CODER_BUF_SIZE ((size_t)1 << 17) - -unsigned Xz_ReadVarInt(const Byte *p, size_t maxSize, UInt64 *value) -{ - unsigned i, limit; - *value = 0; - limit = (maxSize > 9) ? 9 : (unsigned)maxSize; - - for (i = 0; i < limit;) - { - Byte b = p[i]; - *value |= (UInt64)(b & 0x7F) << (7 * i++); - if ((b & 0x80) == 0) - return (b == 0 && i != 1) ? 0 : i; - } - return 0; -} - -/* ---------- BraState ---------- */ - -#define BRA_BUF_SIZE (1 << 14) - -typedef struct -{ - size_t bufPos; - size_t bufConv; - size_t bufTotal; - - int encodeMode; - - UInt32 methodId; - UInt32 delta; - UInt32 ip; - UInt32 x86State; - Byte deltaState[DELTA_STATE_SIZE]; - - Byte buf[BRA_BUF_SIZE]; -} CBraState; - -static void BraState_Free(void *pp, ISzAllocPtr alloc) -{ - ISzAlloc_Free(alloc, pp); -} - -static SRes BraState_SetProps(void *pp, const Byte *props, size_t propSize, ISzAllocPtr alloc) -{ - CBraState *p = ((CBraState *)pp); - UNUSED_VAR(alloc); - p->ip = 0; - if (p->methodId == XZ_ID_Delta) - { - if (propSize != 1) - return SZ_ERROR_UNSUPPORTED; - p->delta = (unsigned)props[0] + 1; - } - else - { - if (propSize == 4) - { - UInt32 v = GetUi32(props); - switch (p->methodId) - { - case XZ_ID_PPC: - case XZ_ID_ARM: - case XZ_ID_SPARC: - if ((v & 3) != 0) - return SZ_ERROR_UNSUPPORTED; - break; - case XZ_ID_ARMT: - if ((v & 1) != 0) - return SZ_ERROR_UNSUPPORTED; - break; - case XZ_ID_IA64: - if ((v & 0xF) != 0) - return SZ_ERROR_UNSUPPORTED; - break; - } - p->ip = v; - } - else if (propSize != 0) - return SZ_ERROR_UNSUPPORTED; - } - return SZ_OK; -} - -static void BraState_Init(void *pp) -{ - CBraState *p = ((CBraState *)pp); - p->bufPos = p->bufConv = p->bufTotal = 0; - x86_Convert_Init(p->x86State); - if (p->methodId == XZ_ID_Delta) - Delta_Init(p->deltaState); -} - - -#define CASE_BRA_CONV(isa) case XZ_ID_ ## isa: size = isa ## _Convert(data, size, p->ip, p->encodeMode); break; - -static SizeT BraState_Filter(void *pp, Byte *data, SizeT size) -{ - CBraState *p = ((CBraState *)pp); - switch (p->methodId) - { - case XZ_ID_Delta: - if (p->encodeMode) - Delta_Encode(p->deltaState, p->delta, data, size); - else - Delta_Decode(p->deltaState, p->delta, data, size); - break; - case XZ_ID_X86: - size = x86_Convert(data, size, p->ip, &p->x86State, p->encodeMode); - break; - CASE_BRA_CONV(PPC) - CASE_BRA_CONV(IA64) - CASE_BRA_CONV(ARM) - CASE_BRA_CONV(ARMT) - CASE_BRA_CONV(SPARC) - } - p->ip += (UInt32)size; - return size; -} - - -static SRes BraState_Code2(void *pp, - Byte *dest, SizeT *destLen, - const Byte *src, SizeT *srcLen, int srcWasFinished, - ECoderFinishMode finishMode, - // int *wasFinished - ECoderStatus *status) -{ - CBraState *p = ((CBraState *)pp); - SizeT destRem = *destLen; - SizeT srcRem = *srcLen; - UNUSED_VAR(finishMode); - - *destLen = 0; - *srcLen = 0; - // *wasFinished = False; - *status = CODER_STATUS_NOT_FINISHED; - - while (destRem > 0) - { - if (p->bufPos != p->bufConv) - { - size_t size = p->bufConv - p->bufPos; - if (size > destRem) - size = destRem; - memcpy(dest, p->buf + p->bufPos, size); - p->bufPos += size; - *destLen += size; - dest += size; - destRem -= size; - continue; - } - - p->bufTotal -= p->bufPos; - memmove(p->buf, p->buf + p->bufPos, p->bufTotal); - p->bufPos = 0; - p->bufConv = 0; - { - size_t size = BRA_BUF_SIZE - p->bufTotal; - if (size > srcRem) - size = srcRem; - memcpy(p->buf + p->bufTotal, src, size); - *srcLen += size; - src += size; - srcRem -= size; - p->bufTotal += size; - } - if (p->bufTotal == 0) - break; - - p->bufConv = BraState_Filter(pp, p->buf, p->bufTotal); - - if (p->bufConv == 0) - { - if (!srcWasFinished) - break; - p->bufConv = p->bufTotal; - } - } - - if (p->bufTotal == p->bufPos && srcRem == 0 && srcWasFinished) - { - *status = CODER_STATUS_FINISHED_WITH_MARK; - // *wasFinished = 1; - } - - return SZ_OK; -} - - -SRes BraState_SetFromMethod(IStateCoder *p, UInt64 id, int encodeMode, ISzAllocPtr alloc); -SRes BraState_SetFromMethod(IStateCoder *p, UInt64 id, int encodeMode, ISzAllocPtr alloc) -{ - CBraState *decoder; - if (id < XZ_ID_Delta || id > XZ_ID_SPARC) - return SZ_ERROR_UNSUPPORTED; - decoder = (CBraState *)p->p; - if (!decoder) - { - decoder = (CBraState *)ISzAlloc_Alloc(alloc, sizeof(CBraState)); - if (!decoder) - return SZ_ERROR_MEM; - p->p = decoder; - p->Free = BraState_Free; - p->SetProps = BraState_SetProps; - p->Init = BraState_Init; - p->Code2 = BraState_Code2; - p->Filter = BraState_Filter; - } - decoder->methodId = (UInt32)id; - decoder->encodeMode = encodeMode; - return SZ_OK; -} - - - -/* ---------- SbState ---------- */ - -#ifdef USE_SUBBLOCK - -static void SbState_Free(void *pp, ISzAllocPtr alloc) -{ - CSbDec *p = (CSbDec *)pp; - SbDec_Free(p); - ISzAlloc_Free(alloc, pp); -} - -static SRes SbState_SetProps(void *pp, const Byte *props, size_t propSize, ISzAllocPtr alloc) -{ - UNUSED_VAR(pp); - UNUSED_VAR(props); - UNUSED_VAR(alloc); - return (propSize == 0) ? SZ_OK : SZ_ERROR_UNSUPPORTED; -} - -static void SbState_Init(void *pp) -{ - SbDec_Init((CSbDec *)pp); -} - -static SRes SbState_Code2(void *pp, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, - int srcWasFinished, ECoderFinishMode finishMode, - // int *wasFinished - ECoderStatus *status) -{ - CSbDec *p = (CSbDec *)pp; - SRes res; - UNUSED_VAR(srcWasFinished); - p->dest = dest; - p->destLen = *destLen; - p->src = src; - p->srcLen = *srcLen; - p->finish = finishMode; /* change it */ - res = SbDec_Decode((CSbDec *)pp); - *destLen -= p->destLen; - *srcLen -= p->srcLen; - // *wasFinished = (*destLen == 0 && *srcLen == 0); /* change it */ - *status = (*destLen == 0 && *srcLen == 0) ? - CODER_STATUS_FINISHED_WITH_MARK : - CODER_STATUS_NOT_FINISHED; - return res; -} - -static SRes SbState_SetFromMethod(IStateCoder *p, ISzAllocPtr alloc) -{ - CSbDec *decoder = (CSbDec *)p->p; - if (!decoder) - { - decoder = (CSbDec *)ISzAlloc_Alloc(alloc, sizeof(CSbDec)); - if (!decoder) - return SZ_ERROR_MEM; - p->p = decoder; - p->Free = SbState_Free; - p->SetProps = SbState_SetProps; - p->Init = SbState_Init; - p->Code2 = SbState_Code2; - p->Filter = NULL; - } - SbDec_Construct(decoder); - SbDec_SetAlloc(decoder, alloc); - return SZ_OK; -} - -#endif - - - -/* ---------- Lzma2 ---------- */ - -typedef struct -{ - CLzma2Dec decoder; - BoolInt outBufMode; -} CLzma2Dec_Spec; - - -static void Lzma2State_Free(void *pp, ISzAllocPtr alloc) -{ - CLzma2Dec_Spec *p = (CLzma2Dec_Spec *)pp; - if (p->outBufMode) - Lzma2Dec_FreeProbs(&p->decoder, alloc); - else - Lzma2Dec_Free(&p->decoder, alloc); - ISzAlloc_Free(alloc, pp); -} - -static SRes Lzma2State_SetProps(void *pp, const Byte *props, size_t propSize, ISzAllocPtr alloc) -{ - if (propSize != 1) - return SZ_ERROR_UNSUPPORTED; - { - CLzma2Dec_Spec *p = (CLzma2Dec_Spec *)pp; - if (p->outBufMode) - return Lzma2Dec_AllocateProbs(&p->decoder, props[0], alloc); - else - return Lzma2Dec_Allocate(&p->decoder, props[0], alloc); - } -} - -static void Lzma2State_Init(void *pp) -{ - Lzma2Dec_Init(&((CLzma2Dec_Spec *)pp)->decoder); -} - - -/* - if (outBufMode), then (dest) is not used. Use NULL. - Data is unpacked to (spec->decoder.decoder.dic) output buffer. -*/ - -static SRes Lzma2State_Code2(void *pp, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, - int srcWasFinished, ECoderFinishMode finishMode, - // int *wasFinished, - ECoderStatus *status) -{ - CLzma2Dec_Spec *spec = (CLzma2Dec_Spec *)pp; - ELzmaStatus status2; - /* ELzmaFinishMode fm = (finishMode == LZMA_FINISH_ANY) ? LZMA_FINISH_ANY : LZMA_FINISH_END; */ - SRes res; - UNUSED_VAR(srcWasFinished); - if (spec->outBufMode) - { - SizeT dicPos = spec->decoder.decoder.dicPos; - SizeT dicLimit = dicPos + *destLen; - res = Lzma2Dec_DecodeToDic(&spec->decoder, dicLimit, src, srcLen, (ELzmaFinishMode)finishMode, &status2); - *destLen = spec->decoder.decoder.dicPos - dicPos; - } - else - res = Lzma2Dec_DecodeToBuf(&spec->decoder, dest, destLen, src, srcLen, (ELzmaFinishMode)finishMode, &status2); - // *wasFinished = (status2 == LZMA_STATUS_FINISHED_WITH_MARK); - // ECoderStatus values are identical to ELzmaStatus values of LZMA2 decoder - *status = (ECoderStatus)status2; - return res; -} - - -static SRes Lzma2State_SetFromMethod(IStateCoder *p, Byte *outBuf, size_t outBufSize, ISzAllocPtr alloc) -{ - CLzma2Dec_Spec *spec = (CLzma2Dec_Spec *)p->p; - if (!spec) - { - spec = (CLzma2Dec_Spec *)ISzAlloc_Alloc(alloc, sizeof(CLzma2Dec_Spec)); - if (!spec) - return SZ_ERROR_MEM; - p->p = spec; - p->Free = Lzma2State_Free; - p->SetProps = Lzma2State_SetProps; - p->Init = Lzma2State_Init; - p->Code2 = Lzma2State_Code2; - p->Filter = NULL; - Lzma2Dec_Construct(&spec->decoder); - } - spec->outBufMode = False; - if (outBuf) - { - spec->outBufMode = True; - spec->decoder.decoder.dic = outBuf; - spec->decoder.decoder.dicBufSize = outBufSize; - } - return SZ_OK; -} - - -static SRes Lzma2State_ResetOutBuf(IStateCoder *p, Byte *outBuf, size_t outBufSize) -{ - CLzma2Dec_Spec *spec = (CLzma2Dec_Spec *)p->p; - if ((spec->outBufMode && !outBuf) || (!spec->outBufMode && outBuf)) - return SZ_ERROR_FAIL; - if (outBuf) - { - spec->decoder.decoder.dic = outBuf; - spec->decoder.decoder.dicBufSize = outBufSize; - } - return SZ_OK; -} - - - -static void MixCoder_Construct(CMixCoder *p, ISzAllocPtr alloc) -{ - unsigned i; - p->alloc = alloc; - p->buf = NULL; - p->numCoders = 0; - - p->outBufSize = 0; - p->outBuf = NULL; - // p->SingleBufMode = False; - - for (i = 0; i < MIXCODER_NUM_FILTERS_MAX; i++) - p->coders[i].p = NULL; -} - - -static void MixCoder_Free(CMixCoder *p) -{ - unsigned i; - p->numCoders = 0; - for (i = 0; i < MIXCODER_NUM_FILTERS_MAX; i++) - { - IStateCoder *sc = &p->coders[i]; - if (sc->p) - { - sc->Free(sc->p, p->alloc); - sc->p = NULL; - } - } - if (p->buf) - { - ISzAlloc_Free(p->alloc, p->buf); - p->buf = NULL; /* 9.31: the BUG was fixed */ - } -} - -static void MixCoder_Init(CMixCoder *p) -{ - unsigned i; - for (i = 0; i < MIXCODER_NUM_FILTERS_MAX - 1; i++) - { - p->size[i] = 0; - p->pos[i] = 0; - p->finished[i] = 0; - } - for (i = 0; i < p->numCoders; i++) - { - IStateCoder *coder = &p->coders[i]; - coder->Init(coder->p); - p->results[i] = SZ_OK; - } - p->outWritten = 0; - p->wasFinished = False; - p->res = SZ_OK; - p->status = CODER_STATUS_NOT_SPECIFIED; -} - - -static SRes MixCoder_SetFromMethod(CMixCoder *p, unsigned coderIndex, UInt64 methodId, Byte *outBuf, size_t outBufSize) -{ - IStateCoder *sc = &p->coders[coderIndex]; - p->ids[coderIndex] = methodId; - switch (methodId) - { - case XZ_ID_LZMA2: return Lzma2State_SetFromMethod(sc, outBuf, outBufSize, p->alloc); - #ifdef USE_SUBBLOCK - case XZ_ID_Subblock: return SbState_SetFromMethod(sc, p->alloc); - #endif - } - if (coderIndex == 0) - return SZ_ERROR_UNSUPPORTED; - return BraState_SetFromMethod(sc, methodId, 0, p->alloc); -} - - -static SRes MixCoder_ResetFromMethod(CMixCoder *p, unsigned coderIndex, UInt64 methodId, Byte *outBuf, size_t outBufSize) -{ - IStateCoder *sc = &p->coders[coderIndex]; - switch (methodId) - { - case XZ_ID_LZMA2: return Lzma2State_ResetOutBuf(sc, outBuf, outBufSize); - } - return SZ_ERROR_UNSUPPORTED; -} - - - -/* - if (destFinish) - then unpack data block is finished at (*destLen) position, - and we can return data that were not processed by filter - -output (status) can be : - CODER_STATUS_NOT_FINISHED - CODER_STATUS_FINISHED_WITH_MARK - CODER_STATUS_NEEDS_MORE_INPUT - not implemented still -*/ - -static SRes MixCoder_Code(CMixCoder *p, - Byte *dest, SizeT *destLen, int destFinish, - const Byte *src, SizeT *srcLen, int srcWasFinished, - ECoderFinishMode finishMode) -{ - SizeT destLenOrig = *destLen; - SizeT srcLenOrig = *srcLen; - - *destLen = 0; - *srcLen = 0; - - if (p->wasFinished) - return p->res; - - p->status = CODER_STATUS_NOT_FINISHED; - - // if (p->SingleBufMode) - if (p->outBuf) - { - SRes res; - SizeT destLen2, srcLen2; - int wasFinished; - - PRF_STR("------- MixCoder Single ----------"); - - srcLen2 = srcLenOrig; - destLen2 = destLenOrig; - - { - IStateCoder *coder = &p->coders[0]; - res = coder->Code2(coder->p, NULL, &destLen2, src, &srcLen2, srcWasFinished, finishMode, - // &wasFinished, - &p->status); - wasFinished = (p->status == CODER_STATUS_FINISHED_WITH_MARK); - } - - p->res = res; - - /* - if (wasFinished) - p->status = CODER_STATUS_FINISHED_WITH_MARK; - else - { - if (res == SZ_OK) - if (destLen2 != destLenOrig) - p->status = CODER_STATUS_NEEDS_MORE_INPUT; - } - */ - - - *srcLen = srcLen2; - src += srcLen2; - p->outWritten += destLen2; - - if (res != SZ_OK || srcWasFinished || wasFinished) - p->wasFinished = True; - - if (p->numCoders == 1) - *destLen = destLen2; - else if (p->wasFinished) - { - unsigned i; - size_t processed = p->outWritten; - - for (i = 1; i < p->numCoders; i++) - { - IStateCoder *coder = &p->coders[i]; - processed = coder->Filter(coder->p, p->outBuf, processed); - if (wasFinished || (destFinish && p->outWritten == destLenOrig)) - processed = p->outWritten; - PRF_STR_INT("filter", i); - } - *destLen = processed; - } - return res; - } - - PRF_STR("standard mix"); - - if (p->numCoders != 1) - { - if (!p->buf) - { - p->buf = (Byte *)ISzAlloc_Alloc(p->alloc, CODER_BUF_SIZE * (MIXCODER_NUM_FILTERS_MAX - 1)); - if (!p->buf) - return SZ_ERROR_MEM; - } - - finishMode = CODER_FINISH_ANY; - } - - for (;;) - { - BoolInt processed = False; - BoolInt allFinished = True; - SRes resMain = SZ_OK; - unsigned i; - - p->status = CODER_STATUS_NOT_FINISHED; - /* - if (p->numCoders == 1 && *destLen == destLenOrig && finishMode == LZMA_FINISH_ANY) - break; - */ - - for (i = 0; i < p->numCoders; i++) - { - SRes res; - IStateCoder *coder = &p->coders[i]; - Byte *dest2; - SizeT destLen2, srcLen2; // destLen2_Orig; - const Byte *src2; - int srcFinished2; - int encodingWasFinished; - ECoderStatus status2; - - if (i == 0) - { - src2 = src; - srcLen2 = srcLenOrig - *srcLen; - srcFinished2 = srcWasFinished; - } - else - { - size_t k = i - 1; - src2 = p->buf + (CODER_BUF_SIZE * k) + p->pos[k]; - srcLen2 = p->size[k] - p->pos[k]; - srcFinished2 = p->finished[k]; - } - - if (i == p->numCoders - 1) - { - dest2 = dest; - destLen2 = destLenOrig - *destLen; - } - else - { - if (p->pos[i] != p->size[i]) - continue; - dest2 = p->buf + (CODER_BUF_SIZE * i); - destLen2 = CODER_BUF_SIZE; - } - - // destLen2_Orig = destLen2; - - if (p->results[i] != SZ_OK) - { - if (resMain == SZ_OK) - resMain = p->results[i]; - continue; - } - - res = coder->Code2(coder->p, - dest2, &destLen2, - src2, &srcLen2, srcFinished2, - finishMode, - // &encodingWasFinished, - &status2); - - if (res != SZ_OK) - { - p->results[i] = res; - if (resMain == SZ_OK) - resMain = res; - } - - encodingWasFinished = (status2 == CODER_STATUS_FINISHED_WITH_MARK); - - if (!encodingWasFinished) - { - allFinished = False; - if (p->numCoders == 1 && res == SZ_OK) - p->status = status2; - } - - if (i == 0) - { - *srcLen += srcLen2; - src += srcLen2; - } - else - p->pos[(size_t)i - 1] += srcLen2; - - if (i == p->numCoders - 1) - { - *destLen += destLen2; - dest += destLen2; - } - else - { - p->size[i] = destLen2; - p->pos[i] = 0; - p->finished[i] = encodingWasFinished; - } - - if (destLen2 != 0 || srcLen2 != 0) - processed = True; - } - - if (!processed) - { - if (allFinished) - p->status = CODER_STATUS_FINISHED_WITH_MARK; - return resMain; - } - } -} - - -SRes Xz_ParseHeader(CXzStreamFlags *p, const Byte *buf) -{ - *p = (CXzStreamFlags)GetBe16(buf + XZ_SIG_SIZE); - if (CrcCalc(buf + XZ_SIG_SIZE, XZ_STREAM_FLAGS_SIZE) != - GetUi32(buf + XZ_SIG_SIZE + XZ_STREAM_FLAGS_SIZE)) - return SZ_ERROR_NO_ARCHIVE; - return XzFlags_IsSupported(*p) ? SZ_OK : SZ_ERROR_UNSUPPORTED; -} - -static BoolInt Xz_CheckFooter(CXzStreamFlags flags, UInt64 indexSize, const Byte *buf) -{ - return indexSize == (((UInt64)GetUi32(buf + 4) + 1) << 2) - && GetUi32(buf) == CrcCalc(buf + 4, 6) - && flags == GetBe16(buf + 8) - && buf[10] == XZ_FOOTER_SIG_0 - && buf[11] == XZ_FOOTER_SIG_1; -} - -#define READ_VARINT_AND_CHECK(buf, pos, size, res) \ - { unsigned s = Xz_ReadVarInt(buf + pos, size - pos, res); \ - if (s == 0) return SZ_ERROR_ARCHIVE; \ - pos += s; } - - -static BoolInt XzBlock_AreSupportedFilters(const CXzBlock *p) -{ - unsigned numFilters = XzBlock_GetNumFilters(p) - 1; - unsigned i; - { - const CXzFilter *f = &p->filters[numFilters]; - if (f->id != XZ_ID_LZMA2 || f->propsSize != 1 || f->props[0] > 40) - return False; - } - - for (i = 0; i < numFilters; i++) - { - const CXzFilter *f = &p->filters[i]; - if (f->id == XZ_ID_Delta) - { - if (f->propsSize != 1) - return False; - } - else if (f->id < XZ_ID_Delta - || f->id > XZ_ID_SPARC - || (f->propsSize != 0 && f->propsSize != 4)) - return False; - } - return True; -} - - -SRes XzBlock_Parse(CXzBlock *p, const Byte *header) -{ - unsigned pos; - unsigned numFilters, i; - unsigned headerSize = (unsigned)header[0] << 2; - - /* (headerSize != 0) : another code checks */ - - if (CrcCalc(header, headerSize) != GetUi32(header + headerSize)) - return SZ_ERROR_ARCHIVE; - - pos = 1; - p->flags = header[pos++]; - - p->packSize = (UInt64)(Int64)-1; - if (XzBlock_HasPackSize(p)) - { - READ_VARINT_AND_CHECK(header, pos, headerSize, &p->packSize); - if (p->packSize == 0 || p->packSize + headerSize >= (UInt64)1 << 63) - return SZ_ERROR_ARCHIVE; - } - - p->unpackSize = (UInt64)(Int64)-1; - if (XzBlock_HasUnpackSize(p)) - READ_VARINT_AND_CHECK(header, pos, headerSize, &p->unpackSize); - - numFilters = XzBlock_GetNumFilters(p); - for (i = 0; i < numFilters; i++) - { - CXzFilter *filter = p->filters + i; - UInt64 size; - READ_VARINT_AND_CHECK(header, pos, headerSize, &filter->id); - READ_VARINT_AND_CHECK(header, pos, headerSize, &size); - if (size > headerSize - pos || size > XZ_FILTER_PROPS_SIZE_MAX) - return SZ_ERROR_ARCHIVE; - filter->propsSize = (UInt32)size; - memcpy(filter->props, header + pos, (size_t)size); - pos += (unsigned)size; - - #ifdef XZ_DUMP - printf("\nf[%u] = %2X: ", i, (unsigned)filter->id); - { - unsigned i; - for (i = 0; i < size; i++) - printf(" %2X", filter->props[i]); - } - #endif - } - - if (XzBlock_HasUnsupportedFlags(p)) - return SZ_ERROR_UNSUPPORTED; - - while (pos < headerSize) - if (header[pos++] != 0) - return SZ_ERROR_ARCHIVE; - return SZ_OK; -} - - - - -static SRes XzDecMix_Init(CMixCoder *p, const CXzBlock *block, Byte *outBuf, size_t outBufSize) -{ - unsigned i; - BoolInt needReInit = True; - unsigned numFilters = XzBlock_GetNumFilters(block); - - if (numFilters == p->numCoders && ((p->outBuf && outBuf) || (!p->outBuf && !outBuf))) - { - needReInit = False; - for (i = 0; i < numFilters; i++) - if (p->ids[i] != block->filters[numFilters - 1 - i].id) - { - needReInit = True; - break; - } - } - - // p->SingleBufMode = (outBuf != NULL); - p->outBuf = outBuf; - p->outBufSize = outBufSize; - - // p->SingleBufMode = False; - // outBuf = NULL; - - if (needReInit) - { - MixCoder_Free(p); - for (i = 0; i < numFilters; i++) - { - RINOK(MixCoder_SetFromMethod(p, i, block->filters[numFilters - 1 - i].id, outBuf, outBufSize)); - } - p->numCoders = numFilters; - } - else - { - RINOK(MixCoder_ResetFromMethod(p, 0, block->filters[numFilters - 1].id, outBuf, outBufSize)); - } - - for (i = 0; i < numFilters; i++) - { - const CXzFilter *f = &block->filters[numFilters - 1 - i]; - IStateCoder *sc = &p->coders[i]; - RINOK(sc->SetProps(sc->p, f->props, f->propsSize, p->alloc)); - } - - MixCoder_Init(p); - return SZ_OK; -} - - - -void XzUnpacker_Init(CXzUnpacker *p) -{ - p->state = XZ_STATE_STREAM_HEADER; - p->pos = 0; - p->numStartedStreams = 0; - p->numFinishedStreams = 0; - p->numTotalBlocks = 0; - p->padSize = 0; - p->decodeOnlyOneBlock = 0; - - p->parseMode = False; - p->decodeToStreamSignature = False; - - // p->outBuf = NULL; - // p->outBufSize = 0; - p->outDataWritten = 0; -} - - -void XzUnpacker_SetOutBuf(CXzUnpacker *p, Byte *outBuf, size_t outBufSize) -{ - p->outBuf = outBuf; - p->outBufSize = outBufSize; -} - - -void XzUnpacker_Construct(CXzUnpacker *p, ISzAllocPtr alloc) -{ - MixCoder_Construct(&p->decoder, alloc); - p->outBuf = NULL; - p->outBufSize = 0; - XzUnpacker_Init(p); -} - - -void XzUnpacker_Free(CXzUnpacker *p) -{ - MixCoder_Free(&p->decoder); -} - - -void XzUnpacker_PrepareToRandomBlockDecoding(CXzUnpacker *p) -{ - p->indexSize = 0; - p->numBlocks = 0; - Sha256_Init(&p->sha); - p->state = XZ_STATE_BLOCK_HEADER; - p->pos = 0; - p->decodeOnlyOneBlock = 1; -} - - -static void XzUnpacker_UpdateIndex(CXzUnpacker *p, UInt64 packSize, UInt64 unpackSize) -{ - Byte temp[32]; - unsigned num = Xz_WriteVarInt(temp, packSize); - num += Xz_WriteVarInt(temp + num, unpackSize); - Sha256_Update(&p->sha, temp, num); - p->indexSize += num; - p->numBlocks++; -} - - - -SRes XzUnpacker_Code(CXzUnpacker *p, Byte *dest, SizeT *destLen, - const Byte *src, SizeT *srcLen, int srcFinished, - ECoderFinishMode finishMode, ECoderStatus *status) -{ - SizeT destLenOrig = *destLen; - SizeT srcLenOrig = *srcLen; - *destLen = 0; - *srcLen = 0; - *status = CODER_STATUS_NOT_SPECIFIED; - - for (;;) - { - SizeT srcRem; - - if (p->state == XZ_STATE_BLOCK) - { - SizeT destLen2 = destLenOrig - *destLen; - SizeT srcLen2 = srcLenOrig - *srcLen; - SRes res; - - ECoderFinishMode finishMode2 = finishMode; - BoolInt srcFinished2 = srcFinished; - BoolInt destFinish = False; - - if (p->block.packSize != (UInt64)(Int64)-1) - { - UInt64 rem = p->block.packSize - p->packSize; - if (srcLen2 >= rem) - { - srcFinished2 = True; - srcLen2 = (SizeT)rem; - } - if (rem == 0 && p->block.unpackSize == p->unpackSize) - return SZ_ERROR_DATA; - } - - if (p->block.unpackSize != (UInt64)(Int64)-1) - { - UInt64 rem = p->block.unpackSize - p->unpackSize; - if (destLen2 >= rem) - { - destFinish = True; - finishMode2 = CODER_FINISH_END; - destLen2 = (SizeT)rem; - } - } - - /* - if (srcLen2 == 0 && destLen2 == 0) - { - *status = CODER_STATUS_NOT_FINISHED; - return SZ_OK; - } - */ - - { - res = MixCoder_Code(&p->decoder, - (p->outBuf ? NULL : dest), &destLen2, destFinish, - src, &srcLen2, srcFinished2, - finishMode2); - - *status = p->decoder.status; - XzCheck_Update(&p->check, (p->outBuf ? p->outBuf + p->outDataWritten : dest), destLen2); - if (!p->outBuf) - dest += destLen2; - p->outDataWritten += destLen2; - } - - (*srcLen) += srcLen2; - src += srcLen2; - p->packSize += srcLen2; - (*destLen) += destLen2; - p->unpackSize += destLen2; - - RINOK(res); - - if (*status != CODER_STATUS_FINISHED_WITH_MARK) - { - if (p->block.packSize == p->packSize - && *status == CODER_STATUS_NEEDS_MORE_INPUT) - { - PRF_STR("CODER_STATUS_NEEDS_MORE_INPUT"); - *status = CODER_STATUS_NOT_SPECIFIED; - return SZ_ERROR_DATA; - } - - return SZ_OK; - } - { - XzUnpacker_UpdateIndex(p, XzUnpacker_GetPackSizeForIndex(p), p->unpackSize); - p->state = XZ_STATE_BLOCK_FOOTER; - p->pos = 0; - p->alignPos = 0; - *status = CODER_STATUS_NOT_SPECIFIED; - - if ((p->block.packSize != (UInt64)(Int64)-1 && p->block.packSize != p->packSize) - || (p->block.unpackSize != (UInt64)(Int64)-1 && p->block.unpackSize != p->unpackSize)) - { - PRF_STR("ERROR: block.size mismatch"); - return SZ_ERROR_DATA; - } - } - // continue; - } - - srcRem = srcLenOrig - *srcLen; - - // XZ_STATE_BLOCK_FOOTER can transit to XZ_STATE_BLOCK_HEADER without input bytes - if (srcRem == 0 && p->state != XZ_STATE_BLOCK_FOOTER) - { - *status = CODER_STATUS_NEEDS_MORE_INPUT; - return SZ_OK; - } - - switch (p->state) - { - case XZ_STATE_STREAM_HEADER: - { - if (p->pos < XZ_STREAM_HEADER_SIZE) - { - if (p->pos < XZ_SIG_SIZE && *src != XZ_SIG[p->pos]) - return SZ_ERROR_NO_ARCHIVE; - if (p->decodeToStreamSignature) - return SZ_OK; - p->buf[p->pos++] = *src++; - (*srcLen)++; - } - else - { - RINOK(Xz_ParseHeader(&p->streamFlags, p->buf)); - p->numStartedStreams++; - p->indexSize = 0; - p->numBlocks = 0; - Sha256_Init(&p->sha); - p->state = XZ_STATE_BLOCK_HEADER; - p->pos = 0; - } - break; - } - - case XZ_STATE_BLOCK_HEADER: - { - if (p->pos == 0) - { - p->buf[p->pos++] = *src++; - (*srcLen)++; - if (p->buf[0] == 0) - { - if (p->decodeOnlyOneBlock) - return SZ_ERROR_DATA; - p->indexPreSize = 1 + Xz_WriteVarInt(p->buf + 1, p->numBlocks); - p->indexPos = p->indexPreSize; - p->indexSize += p->indexPreSize; - Sha256_Final(&p->sha, p->shaDigest); - Sha256_Init(&p->sha); - p->crc = CrcUpdate(CRC_INIT_VAL, p->buf, p->indexPreSize); - p->state = XZ_STATE_STREAM_INDEX; - break; - } - p->blockHeaderSize = ((UInt32)p->buf[0] << 2) + 4; - break; - } - - if (p->pos != p->blockHeaderSize) - { - UInt32 cur = p->blockHeaderSize - p->pos; - if (cur > srcRem) - cur = (UInt32)srcRem; - memcpy(p->buf + p->pos, src, cur); - p->pos += cur; - (*srcLen) += cur; - src += cur; - } - else - { - RINOK(XzBlock_Parse(&p->block, p->buf)); - if (!XzBlock_AreSupportedFilters(&p->block)) - return SZ_ERROR_UNSUPPORTED; - p->numTotalBlocks++; - p->state = XZ_STATE_BLOCK; - p->packSize = 0; - p->unpackSize = 0; - XzCheck_Init(&p->check, XzFlags_GetCheckType(p->streamFlags)); - if (p->parseMode) - { - p->headerParsedOk = True; - return SZ_OK; - } - RINOK(XzDecMix_Init(&p->decoder, &p->block, p->outBuf, p->outBufSize)); - } - break; - } - - case XZ_STATE_BLOCK_FOOTER: - { - if ((((unsigned)p->packSize + p->alignPos) & 3) != 0) - { - if (srcRem == 0) - { - *status = CODER_STATUS_NEEDS_MORE_INPUT; - return SZ_OK; - } - (*srcLen)++; - p->alignPos++; - if (*src++ != 0) - return SZ_ERROR_CRC; - } - else - { - UInt32 checkSize = XzFlags_GetCheckSize(p->streamFlags); - UInt32 cur = checkSize - p->pos; - if (cur != 0) - { - if (srcRem == 0) - { - *status = CODER_STATUS_NEEDS_MORE_INPUT; - return SZ_OK; - } - if (cur > srcRem) - cur = (UInt32)srcRem; - memcpy(p->buf + p->pos, src, cur); - p->pos += cur; - (*srcLen) += cur; - src += cur; - if (checkSize != p->pos) - break; - } - { - Byte digest[XZ_CHECK_SIZE_MAX]; - p->state = XZ_STATE_BLOCK_HEADER; - p->pos = 0; - if (XzCheck_Final(&p->check, digest) && memcmp(digest, p->buf, checkSize) != 0) - return SZ_ERROR_CRC; - if (p->decodeOnlyOneBlock) - { - *status = CODER_STATUS_FINISHED_WITH_MARK; - return SZ_OK; - } - } - } - break; - } - - case XZ_STATE_STREAM_INDEX: - { - if (p->pos < p->indexPreSize) - { - (*srcLen)++; - if (*src++ != p->buf[p->pos++]) - return SZ_ERROR_CRC; - } - else - { - if (p->indexPos < p->indexSize) - { - UInt64 cur = p->indexSize - p->indexPos; - if (srcRem > cur) - srcRem = (SizeT)cur; - p->crc = CrcUpdate(p->crc, src, srcRem); - Sha256_Update(&p->sha, src, srcRem); - (*srcLen) += srcRem; - src += srcRem; - p->indexPos += srcRem; - } - else if ((p->indexPos & 3) != 0) - { - Byte b = *src++; - p->crc = CRC_UPDATE_BYTE(p->crc, b); - (*srcLen)++; - p->indexPos++; - p->indexSize++; - if (b != 0) - return SZ_ERROR_CRC; - } - else - { - Byte digest[SHA256_DIGEST_SIZE]; - p->state = XZ_STATE_STREAM_INDEX_CRC; - p->indexSize += 4; - p->pos = 0; - Sha256_Final(&p->sha, digest); - if (memcmp(digest, p->shaDigest, SHA256_DIGEST_SIZE) != 0) - return SZ_ERROR_CRC; - } - } - break; - } - - case XZ_STATE_STREAM_INDEX_CRC: - { - if (p->pos < 4) - { - (*srcLen)++; - p->buf[p->pos++] = *src++; - } - else - { - const Byte *ptr = p->buf; - p->state = XZ_STATE_STREAM_FOOTER; - p->pos = 0; - if (CRC_GET_DIGEST(p->crc) != GetUi32(ptr)) - return SZ_ERROR_CRC; - } - break; - } - - case XZ_STATE_STREAM_FOOTER: - { - UInt32 cur = XZ_STREAM_FOOTER_SIZE - p->pos; - if (cur > srcRem) - cur = (UInt32)srcRem; - memcpy(p->buf + p->pos, src, cur); - p->pos += cur; - (*srcLen) += cur; - src += cur; - if (p->pos == XZ_STREAM_FOOTER_SIZE) - { - p->state = XZ_STATE_STREAM_PADDING; - p->numFinishedStreams++; - p->padSize = 0; - if (!Xz_CheckFooter(p->streamFlags, p->indexSize, p->buf)) - return SZ_ERROR_CRC; - } - break; - } - - case XZ_STATE_STREAM_PADDING: - { - if (*src != 0) - { - if (((UInt32)p->padSize & 3) != 0) - return SZ_ERROR_NO_ARCHIVE; - p->pos = 0; - p->state = XZ_STATE_STREAM_HEADER; - } - else - { - (*srcLen)++; - src++; - p->padSize++; - } - break; - } - - case XZ_STATE_BLOCK: break; /* to disable GCC warning */ - } - } - /* - if (p->state == XZ_STATE_FINISHED) - *status = CODER_STATUS_FINISHED_WITH_MARK; - return SZ_OK; - */ -} - - -SRes XzUnpacker_CodeFull(CXzUnpacker *p, Byte *dest, SizeT *destLen, - const Byte *src, SizeT *srcLen, - ECoderFinishMode finishMode, ECoderStatus *status) -{ - XzUnpacker_Init(p); - XzUnpacker_SetOutBuf(p, dest, *destLen); - - return XzUnpacker_Code(p, - NULL, destLen, - src, srcLen, True, - finishMode, status); -} - - -BoolInt XzUnpacker_IsBlockFinished(const CXzUnpacker *p) -{ - return (p->state == XZ_STATE_BLOCK_HEADER) && (p->pos == 0); -} - -BoolInt XzUnpacker_IsStreamWasFinished(const CXzUnpacker *p) -{ - return (p->state == XZ_STATE_STREAM_PADDING) && (((UInt32)p->padSize & 3) == 0); -} - -UInt64 XzUnpacker_GetExtraSize(const CXzUnpacker *p) -{ - UInt64 num = 0; - if (p->state == XZ_STATE_STREAM_PADDING) - num = p->padSize; - else if (p->state == XZ_STATE_STREAM_HEADER) - num = p->padSize + p->pos; - return num; -} - - - - - - - - - - - - - - - - - - - - - -#ifndef _7ZIP_ST -#include "MtDec.h" -#endif - - -void XzDecMtProps_Init(CXzDecMtProps *p) -{ - p->inBufSize_ST = 1 << 18; - p->outStep_ST = 1 << 20; - p->ignoreErrors = False; - - #ifndef _7ZIP_ST - p->numThreads = 1; - p->inBufSize_MT = 1 << 18; - p->memUseMax = sizeof(size_t) << 28; - #endif -} - - - -#ifndef _7ZIP_ST - -/* ---------- CXzDecMtThread ---------- */ - -typedef struct -{ - Byte *outBuf; - size_t outBufSize; - size_t outPreSize; - size_t inPreSize; - size_t inPreHeaderSize; - size_t blockPackSize_for_Index; // including block header and checksum. - size_t blockPackTotal; // including stream header, block header and checksum. - size_t inCodeSize; - size_t outCodeSize; - ECoderStatus status; - SRes codeRes; - BoolInt skipMode; - // BoolInt finishedWithMark; - EMtDecParseState parseState; - BoolInt parsing_Truncated; - BoolInt atBlockHeader; - CXzStreamFlags streamFlags; - // UInt64 numFinishedStreams - UInt64 numStreams; - UInt64 numTotalBlocks; - UInt64 numBlocks; - - BoolInt dec_created; - CXzUnpacker dec; - - Byte mtPad[1 << 7]; -} CXzDecMtThread; - -#endif - - -/* ---------- CXzDecMt ---------- */ - -typedef struct -{ - CAlignOffsetAlloc alignOffsetAlloc; - ISzAllocPtr allocMid; - - CXzDecMtProps props; - size_t unpackBlockMaxSize; - - ISeqInStream *inStream; - ISeqOutStream *outStream; - ICompressProgress *progress; - - BoolInt finishMode; - BoolInt outSize_Defined; - UInt64 outSize; - - UInt64 outProcessed; - UInt64 inProcessed; - UInt64 readProcessed; - BoolInt readWasFinished; - SRes readRes; - SRes writeRes; - - Byte *outBuf; - size_t outBufSize; - Byte *inBuf; - size_t inBufSize; - - CXzUnpacker dec; - - ECoderStatus status; - SRes codeRes; - - #ifndef _7ZIP_ST - BoolInt mainDecoderWasCalled; - // int statErrorDefined; - int finishedDecoderIndex; - - // global values that are used in Parse stage - CXzStreamFlags streamFlags; - // UInt64 numFinishedStreams - UInt64 numStreams; - UInt64 numTotalBlocks; - UInt64 numBlocks; - - // UInt64 numBadBlocks; - SRes mainErrorCode; // it's set to error code, if the size Code() output doesn't patch the size from Parsing stage - // it can be = SZ_ERROR_INPUT_EOF - // it can be = SZ_ERROR_DATA, in some another cases - BoolInt isBlockHeaderState_Parse; - BoolInt isBlockHeaderState_Write; - UInt64 outProcessed_Parse; - BoolInt parsing_Truncated; - - BoolInt mtc_WasConstructed; - CMtDec mtc; - CXzDecMtThread coders[MTDEC__THREADS_MAX]; - #endif - -} CXzDecMt; - - - -CXzDecMtHandle XzDecMt_Create(ISzAllocPtr alloc, ISzAllocPtr allocMid) -{ - CXzDecMt *p = (CXzDecMt *)ISzAlloc_Alloc(alloc, sizeof(CXzDecMt)); - if (!p) - return NULL; - - AlignOffsetAlloc_CreateVTable(&p->alignOffsetAlloc); - p->alignOffsetAlloc.baseAlloc = alloc; - p->alignOffsetAlloc.numAlignBits = 7; - p->alignOffsetAlloc.offset = 0; - - p->allocMid = allocMid; - - p->outBuf = NULL; - p->outBufSize = 0; - p->inBuf = NULL; - p->inBufSize = 0; - - XzUnpacker_Construct(&p->dec, &p->alignOffsetAlloc.vt); - - p->unpackBlockMaxSize = 0; - - XzDecMtProps_Init(&p->props); - - #ifndef _7ZIP_ST - p->mtc_WasConstructed = False; - { - unsigned i; - for (i = 0; i < MTDEC__THREADS_MAX; i++) - { - CXzDecMtThread *coder = &p->coders[i]; - coder->dec_created = False; - coder->outBuf = NULL; - coder->outBufSize = 0; - } - } - #endif - - return p; -} - - -#ifndef _7ZIP_ST - -static void XzDecMt_FreeOutBufs(CXzDecMt *p) -{ - unsigned i; - for (i = 0; i < MTDEC__THREADS_MAX; i++) - { - CXzDecMtThread *coder = &p->coders[i]; - if (coder->outBuf) - { - ISzAlloc_Free(p->allocMid, coder->outBuf); - coder->outBuf = NULL; - coder->outBufSize = 0; - } - } - p->unpackBlockMaxSize = 0; -} - -#endif - - - -static void XzDecMt_FreeSt(CXzDecMt *p) -{ - XzUnpacker_Free(&p->dec); - - if (p->outBuf) - { - ISzAlloc_Free(p->allocMid, p->outBuf); - p->outBuf = NULL; - } - p->outBufSize = 0; - - if (p->inBuf) - { - ISzAlloc_Free(p->allocMid, p->inBuf); - p->inBuf = NULL; - } - p->inBufSize = 0; -} - - -void XzDecMt_Destroy(CXzDecMtHandle pp) -{ - CXzDecMt *p = (CXzDecMt *)pp; - - XzDecMt_FreeSt(p); - - #ifndef _7ZIP_ST - - if (p->mtc_WasConstructed) - { - MtDec_Destruct(&p->mtc); - p->mtc_WasConstructed = False; - } - { - unsigned i; - for (i = 0; i < MTDEC__THREADS_MAX; i++) - { - CXzDecMtThread *t = &p->coders[i]; - if (t->dec_created) - { - // we don't need to free dict here - XzUnpacker_Free(&t->dec); - t->dec_created = False; - } - } - } - XzDecMt_FreeOutBufs(p); - - #endif - - ISzAlloc_Free(p->alignOffsetAlloc.baseAlloc, pp); -} - - - -#ifndef _7ZIP_ST - -static void XzDecMt_Callback_Parse(void *obj, unsigned coderIndex, CMtDecCallbackInfo *cc) -{ - CXzDecMt *me = (CXzDecMt *)obj; - CXzDecMtThread *coder = &me->coders[coderIndex]; - size_t srcSize = cc->srcSize; - - cc->srcSize = 0; - cc->outPos = 0; - cc->state = MTDEC_PARSE_CONTINUE; - - cc->canCreateNewThread = True; - - if (cc->startCall) - { - coder->outPreSize = 0; - coder->inPreSize = 0; - coder->inPreHeaderSize = 0; - coder->parseState = MTDEC_PARSE_CONTINUE; - coder->parsing_Truncated = False; - coder->skipMode = False; - coder->codeRes = SZ_OK; - coder->status = CODER_STATUS_NOT_SPECIFIED; - coder->inCodeSize = 0; - coder->outCodeSize = 0; - - coder->numStreams = me->numStreams; - coder->numTotalBlocks = me->numTotalBlocks; - coder->numBlocks = me->numBlocks; - - if (!coder->dec_created) - { - XzUnpacker_Construct(&coder->dec, &me->alignOffsetAlloc.vt); - coder->dec_created = True; - } - - XzUnpacker_Init(&coder->dec); - - if (me->isBlockHeaderState_Parse) - { - coder->dec.streamFlags = me->streamFlags; - coder->atBlockHeader = True; - XzUnpacker_PrepareToRandomBlockDecoding(&coder->dec); - } - else - { - coder->atBlockHeader = False; - me->isBlockHeaderState_Parse = True; - } - - coder->dec.numStartedStreams = me->numStreams; - coder->dec.numTotalBlocks = me->numTotalBlocks; - coder->dec.numBlocks = me->numBlocks; - } - - while (!coder->skipMode) - { - ECoderStatus status; - SRes res; - size_t srcSize2 = srcSize; - size_t destSize = (size_t)0 - 1; - - coder->dec.parseMode = True; - coder->dec.headerParsedOk = False; - - PRF_STR_INT("Parse", srcSize2); - - res = XzUnpacker_Code(&coder->dec, - NULL, &destSize, - cc->src, &srcSize2, cc->srcFinished, - CODER_FINISH_END, &status); - - // PRF(printf(" res = %d, srcSize2 = %d", res, (unsigned)srcSize2)); - - coder->codeRes = res; - coder->status = status; - cc->srcSize += srcSize2; - srcSize -= srcSize2; - coder->inPreHeaderSize += srcSize2; - coder->inPreSize = coder->inPreHeaderSize; - - if (res != SZ_OK) - { - cc->state = - coder->parseState = MTDEC_PARSE_END; - /* - if (res == SZ_ERROR_MEM) - return res; - return SZ_OK; - */ - return; // res; - } - - if (coder->dec.headerParsedOk) - { - const CXzBlock *block = &coder->dec.block; - if (XzBlock_HasUnpackSize(block) - // && block->unpackSize <= me->props.outBlockMax - && XzBlock_HasPackSize(block)) - { - { - if (block->unpackSize * 2 * me->mtc.numStartedThreads > me->props.memUseMax) - { - cc->state = MTDEC_PARSE_OVERFLOW; - return; // SZ_OK; - } - } - { - UInt64 packSize = block->packSize; - UInt64 packSizeAligned = packSize + ((0 - (unsigned)packSize) & 3); - UInt32 checkSize = XzFlags_GetCheckSize(coder->dec.streamFlags); - UInt64 blockPackSum = coder->inPreSize + packSizeAligned + checkSize; - // if (blockPackSum <= me->props.inBlockMax) - // unpackBlockMaxSize - { - coder->blockPackSize_for_Index = (size_t)(coder->dec.blockHeaderSize + packSize + checkSize); - coder->blockPackTotal = (size_t)blockPackSum; - coder->outPreSize = (size_t)block->unpackSize; - coder->streamFlags = coder->dec.streamFlags; - me->streamFlags = coder->dec.streamFlags; - coder->skipMode = True; - break; - } - } - } - } - else - // if (coder->inPreSize <= me->props.inBlockMax) - { - if (!cc->srcFinished) - return; // SZ_OK; - cc->state = - coder->parseState = MTDEC_PARSE_END; - return; // SZ_OK; - } - cc->state = MTDEC_PARSE_OVERFLOW; - return; // SZ_OK; - } - - // ---------- skipMode ---------- - { - UInt64 rem = coder->blockPackTotal - coder->inPreSize; - size_t cur = srcSize; - if (cur > rem) - cur = (size_t)rem; - cc->srcSize += cur; - coder->inPreSize += cur; - srcSize -= cur; - - if (coder->inPreSize == coder->blockPackTotal) - { - if (srcSize == 0) - { - if (!cc->srcFinished) - return; // SZ_OK; - cc->state = MTDEC_PARSE_END; - } - else if ((cc->src)[cc->srcSize] == 0) // we check control byte of next block - cc->state = MTDEC_PARSE_END; - else - { - cc->state = MTDEC_PARSE_NEW; - - { - size_t blockMax = me->unpackBlockMaxSize; - if (blockMax < coder->outPreSize) - blockMax = coder->outPreSize; - { - UInt64 required = (UInt64)blockMax * (me->mtc.numStartedThreads + 1) * 2; - if (me->props.memUseMax < required) - cc->canCreateNewThread = False; - } - } - - if (me->outSize_Defined) - { - // next block can be zero size - const UInt64 rem2 = me->outSize - me->outProcessed_Parse; - if (rem2 < coder->outPreSize) - { - coder->parsing_Truncated = True; - cc->state = MTDEC_PARSE_END; - } - me->outProcessed_Parse += coder->outPreSize; - } - } - } - else if (cc->srcFinished) - cc->state = MTDEC_PARSE_END; - else - return; // SZ_OK; - - coder->parseState = cc->state; - cc->outPos = coder->outPreSize; - - me->numStreams = coder->dec.numStartedStreams; - me->numTotalBlocks = coder->dec.numTotalBlocks; - me->numBlocks = coder->dec.numBlocks + 1; - return; // SZ_OK; - } -} - - -static SRes XzDecMt_Callback_PreCode(void *pp, unsigned coderIndex) -{ - CXzDecMt *me = (CXzDecMt *)pp; - CXzDecMtThread *coder = &me->coders[coderIndex]; - Byte *dest; - - if (!coder->dec.headerParsedOk) - return SZ_OK; - - dest = coder->outBuf; - - if (!dest || coder->outBufSize < coder->outPreSize) - { - if (dest) - { - ISzAlloc_Free(me->allocMid, dest); - coder->outBuf = NULL; - coder->outBufSize = 0; - } - { - size_t outPreSize = coder->outPreSize; - if (outPreSize == 0) - outPreSize = 1; - dest = (Byte *)ISzAlloc_Alloc(me->allocMid, outPreSize); - } - if (!dest) - return SZ_ERROR_MEM; - coder->outBuf = dest; - coder->outBufSize = coder->outPreSize; - - if (coder->outBufSize > me->unpackBlockMaxSize) - me->unpackBlockMaxSize = coder->outBufSize; - } - - // return SZ_ERROR_MEM; - - XzUnpacker_SetOutBuf(&coder->dec, coder->outBuf, coder->outBufSize); - - { - SRes res = XzDecMix_Init(&coder->dec.decoder, &coder->dec.block, coder->outBuf, coder->outBufSize); - // res = SZ_ERROR_UNSUPPORTED; // to test - coder->codeRes = res; - if (res != SZ_OK) - { - // if (res == SZ_ERROR_MEM) return res; - if (me->props.ignoreErrors && res != SZ_ERROR_MEM) - return SZ_OK; - return res; - } - } - - return SZ_OK; -} - - -static SRes XzDecMt_Callback_Code(void *pp, unsigned coderIndex, - const Byte *src, size_t srcSize, int srcFinished, - // int finished, int blockFinished, - UInt64 *inCodePos, UInt64 *outCodePos, int *stop) -{ - CXzDecMt *me = (CXzDecMt *)pp; - CXzDecMtThread *coder = &me->coders[coderIndex]; - - *inCodePos = coder->inCodeSize; - *outCodePos = coder->outCodeSize; - *stop = True; - - if (srcSize > coder->inPreSize - coder->inCodeSize) - return SZ_ERROR_FAIL; - - if (coder->inCodeSize < coder->inPreHeaderSize) - { - size_t step = coder->inPreHeaderSize - coder->inCodeSize; - if (step > srcSize) - step = srcSize; - src += step; - srcSize -= step; - coder->inCodeSize += step; - *inCodePos = coder->inCodeSize; - if (coder->inCodeSize < coder->inPreHeaderSize) - { - *stop = False; - return SZ_OK; - } - } - - if (!coder->dec.headerParsedOk) - return SZ_OK; - if (!coder->outBuf) - return SZ_OK; - - if (coder->codeRes == SZ_OK) - { - ECoderStatus status; - SRes res; - size_t srcProcessed = srcSize; - size_t outSizeCur = coder->outPreSize - coder->dec.outDataWritten; - - // PRF(printf("\nCallback_Code: Code %d %d\n", (unsigned)srcSize, (unsigned)outSizeCur)); - - res = XzUnpacker_Code(&coder->dec, - NULL, &outSizeCur, - src, &srcProcessed, srcFinished, - // coder->finishedWithMark ? CODER_FINISH_END : CODER_FINISH_ANY, - CODER_FINISH_END, - &status); - - // PRF(printf(" res = %d, srcSize2 = %d, outSizeCur = %d", res, (unsigned)srcProcessed, (unsigned)outSizeCur)); - - coder->codeRes = res; - coder->status = status; - coder->inCodeSize += srcProcessed; - coder->outCodeSize = coder->dec.outDataWritten; - *inCodePos = coder->inCodeSize; - *outCodePos = coder->outCodeSize; - - if (res == SZ_OK) - { - if (srcProcessed == srcSize) - *stop = False; - return SZ_OK; - } - } - - if (me->props.ignoreErrors && coder->codeRes != SZ_ERROR_MEM) - { - *inCodePos = coder->inPreSize; - *outCodePos = coder->outPreSize; - return SZ_OK; - } - return coder->codeRes; -} - - -#define XZDECMT_STREAM_WRITE_STEP (1 << 24) - -static SRes XzDecMt_Callback_Write(void *pp, unsigned coderIndex, - BoolInt needWriteToStream, - const Byte *src, size_t srcSize, BoolInt isCross, - // int srcFinished, - BoolInt *needContinue, - BoolInt *canRecode) -{ - CXzDecMt *me = (CXzDecMt *)pp; - const CXzDecMtThread *coder = &me->coders[coderIndex]; - - // PRF(printf("\nWrite processed = %d srcSize = %d\n", (unsigned)me->mtc.inProcessed, (unsigned)srcSize)); - - *needContinue = False; - *canRecode = True; - - if (!needWriteToStream) - return SZ_OK; - - if (!coder->dec.headerParsedOk || !coder->outBuf) - { - if (me->finishedDecoderIndex < 0) - me->finishedDecoderIndex = (int)coderIndex; - return SZ_OK; - } - - if (me->finishedDecoderIndex >= 0) - return SZ_OK; - - me->mtc.inProcessed += coder->inCodeSize; - - *canRecode = False; - - { - SRes res; - size_t size = coder->outCodeSize; - Byte *data = coder->outBuf; - - // we use in me->dec: sha, numBlocks, indexSize - - if (!me->isBlockHeaderState_Write) - { - XzUnpacker_PrepareToRandomBlockDecoding(&me->dec); - me->dec.decodeOnlyOneBlock = False; - me->dec.numStartedStreams = coder->dec.numStartedStreams; - me->dec.streamFlags = coder->streamFlags; - - me->isBlockHeaderState_Write = True; - } - - me->dec.numTotalBlocks = coder->dec.numTotalBlocks; - XzUnpacker_UpdateIndex(&me->dec, coder->blockPackSize_for_Index, coder->outPreSize); - - if (coder->outPreSize != size) - { - if (me->props.ignoreErrors) - { - memset(data + size, 0, coder->outPreSize - size); - size = coder->outPreSize; - } - // me->numBadBlocks++; - if (me->mainErrorCode == SZ_OK) - { - if ((int)coder->status == LZMA_STATUS_NEEDS_MORE_INPUT) - me->mainErrorCode = SZ_ERROR_INPUT_EOF; - else - me->mainErrorCode = SZ_ERROR_DATA; - } - } - - if (me->writeRes != SZ_OK) - return me->writeRes; - - res = SZ_OK; - { - if (me->outSize_Defined) - { - const UInt64 rem = me->outSize - me->outProcessed; - if (size > rem) - size = (SizeT)rem; - } - - for (;;) - { - size_t cur = size; - size_t written; - if (cur > XZDECMT_STREAM_WRITE_STEP) - cur = XZDECMT_STREAM_WRITE_STEP; - - written = ISeqOutStream_Write(me->outStream, data, cur); - - // PRF(printf("\nWritten ask = %d written = %d\n", (unsigned)cur, (unsigned)written)); - - me->outProcessed += written; - if (written != cur) - { - me->writeRes = SZ_ERROR_WRITE; - res = me->writeRes; - break; - } - data += cur; - size -= cur; - // PRF_STR_INT("Written size =", size); - if (size == 0) - break; - res = MtProgress_ProgressAdd(&me->mtc.mtProgress, 0, 0); - if (res != SZ_OK) - break; - } - } - - if (coder->codeRes != SZ_OK) - if (!me->props.ignoreErrors) - { - me->finishedDecoderIndex = (int)coderIndex; - return res; - } - - RINOK(res); - - if (coder->inPreSize != coder->inCodeSize - || coder->blockPackTotal != coder->inCodeSize) - { - me->finishedDecoderIndex = (int)coderIndex; - return SZ_OK; - } - - if (coder->parseState != MTDEC_PARSE_END) - { - *needContinue = True; - return SZ_OK; - } - } - - // (coder->state == MTDEC_PARSE_END) means that there are no other working threads - // so we can use mtc variables without lock - - PRF_STR_INT("Write MTDEC_PARSE_END", me->mtc.inProcessed); - - me->mtc.mtProgress.totalInSize = me->mtc.inProcessed; - { - CXzUnpacker *dec = &me->dec; - - PRF_STR_INT("PostSingle", srcSize); - - { - size_t srcProcessed = srcSize; - ECoderStatus status; - size_t outSizeCur = 0; - SRes res; - - // dec->decodeOnlyOneBlock = False; - dec->decodeToStreamSignature = True; - - me->mainDecoderWasCalled = True; - - if (coder->parsing_Truncated) - { - me->parsing_Truncated = True; - return SZ_OK; - } - - /* - We have processed all xz-blocks of stream, - And xz unpacker is at XZ_STATE_BLOCK_HEADER state, where - (src) is a pointer to xz-Index structure. - We finish reading of current xz-Stream, including Zero padding after xz-Stream. - We exit, if we reach extra byte (first byte of new-Stream or another data). - But we don't update input stream pointer for that new extra byte. - If extra byte is not correct first byte of xz-signature, - we have SZ_ERROR_NO_ARCHIVE error here. - */ - - res = XzUnpacker_Code(dec, - NULL, &outSizeCur, - src, &srcProcessed, - me->mtc.readWasFinished, // srcFinished - CODER_FINISH_END, // CODER_FINISH_ANY, - &status); - - // res = SZ_ERROR_ARCHIVE; // for failure test - - me->status = status; - me->codeRes = res; - - if (isCross) - me->mtc.crossStart += srcProcessed; - - me->mtc.inProcessed += srcProcessed; - me->mtc.mtProgress.totalInSize = me->mtc.inProcessed; - - srcSize -= srcProcessed; - src += srcProcessed; - - if (res != SZ_OK) - { - return SZ_OK; - // return res; - } - - if (dec->state == XZ_STATE_STREAM_HEADER) - { - *needContinue = True; - me->isBlockHeaderState_Parse = False; - me->isBlockHeaderState_Write = False; - - if (!isCross) - { - Byte *crossBuf = MtDec_GetCrossBuff(&me->mtc); - if (!crossBuf) - return SZ_ERROR_MEM; - if (srcSize != 0) - memcpy(crossBuf, src, srcSize); - me->mtc.crossStart = 0; - me->mtc.crossEnd = srcSize; - } - - PRF_STR_INT("XZ_STATE_STREAM_HEADER crossEnd = ", (unsigned)me->mtc.crossEnd); - - return SZ_OK; - } - - if (status != CODER_STATUS_NEEDS_MORE_INPUT || srcSize != 0) - { - return SZ_ERROR_FAIL; - } - - if (me->mtc.readWasFinished) - { - return SZ_OK; - } - } - - { - size_t inPos; - size_t inLim; - // const Byte *inData; - UInt64 inProgressPrev = me->mtc.inProcessed; - - // XzDecMt_Prepare_InBuf_ST(p); - Byte *crossBuf = MtDec_GetCrossBuff(&me->mtc); - if (!crossBuf) - return SZ_ERROR_MEM; - - inPos = 0; - inLim = 0; - - // inData = crossBuf; - - for (;;) - { - SizeT inProcessed; - SizeT outProcessed; - ECoderStatus status; - SRes res; - - if (inPos == inLim) - { - if (!me->mtc.readWasFinished) - { - inPos = 0; - inLim = me->mtc.inBufSize; - me->mtc.readRes = ISeqInStream_Read(me->inStream, (void *)crossBuf, &inLim); - me->mtc.readProcessed += inLim; - if (inLim == 0 || me->mtc.readRes != SZ_OK) - me->mtc.readWasFinished = True; - } - } - - inProcessed = inLim - inPos; - outProcessed = 0; - - res = XzUnpacker_Code(dec, - NULL, &outProcessed, - crossBuf + inPos, &inProcessed, - (inProcessed == 0), // srcFinished - CODER_FINISH_END, &status); - - me->codeRes = res; - me->status = status; - inPos += inProcessed; - me->mtc.inProcessed += inProcessed; - me->mtc.mtProgress.totalInSize = me->mtc.inProcessed; - - if (res != SZ_OK) - { - return SZ_OK; - // return res; - } - - if (dec->state == XZ_STATE_STREAM_HEADER) - { - *needContinue = True; - me->mtc.crossStart = inPos; - me->mtc.crossEnd = inLim; - me->isBlockHeaderState_Parse = False; - me->isBlockHeaderState_Write = False; - return SZ_OK; - } - - if (status != CODER_STATUS_NEEDS_MORE_INPUT) - return SZ_ERROR_FAIL; - - if (me->mtc.progress) - { - UInt64 inDelta = me->mtc.inProcessed - inProgressPrev; - if (inDelta >= (1 << 22)) - { - RINOK(MtProgress_Progress_ST(&me->mtc.mtProgress)); - inProgressPrev = me->mtc.inProcessed; - } - } - if (me->mtc.readWasFinished) - return SZ_OK; - } - } - } -} - - -#endif - - - -void XzStatInfo_Clear(CXzStatInfo *p) -{ - p->InSize = 0; - p->OutSize = 0; - - p->NumStreams = 0; - p->NumBlocks = 0; - - p->UnpackSize_Defined = False; - - p->NumStreams_Defined = False; - p->NumBlocks_Defined = False; - - p->DataAfterEnd = False; - p->DecodingTruncated = False; - - p->DecodeRes = SZ_OK; - p->ReadRes = SZ_OK; - p->ProgressRes = SZ_OK; - - p->CombinedRes = SZ_OK; - p->CombinedRes_Type = SZ_OK; -} - - - -/* - XzDecMt_Decode_ST() can return SZ_OK or the following errors - - SZ_ERROR_MEM for memory allocation error - - error from XzUnpacker_Code() function - - SZ_ERROR_WRITE for ISeqOutStream::Write(). stat->CombinedRes_Type = SZ_ERROR_WRITE in that case - - ICompressProgress::Progress() error, stat->CombinedRes_Type = SZ_ERROR_PROGRESS. - But XzDecMt_Decode_ST() doesn't return ISeqInStream::Read() errors. - ISeqInStream::Read() result is set to p->readRes. - also it can set stat->CombinedRes_Type to SZ_ERROR_WRITE or SZ_ERROR_PROGRESS. -*/ - -static SRes XzDecMt_Decode_ST(CXzDecMt *p - #ifndef _7ZIP_ST - , BoolInt tMode - #endif - , CXzStatInfo *stat) -{ - size_t outPos; - size_t inPos, inLim; - const Byte *inData; - UInt64 inPrev, outPrev; - - CXzUnpacker *dec; - - #ifndef _7ZIP_ST - if (tMode) - { - XzDecMt_FreeOutBufs(p); - tMode = MtDec_PrepareRead(&p->mtc); - } - #endif - - if (!p->outBuf || p->outBufSize != p->props.outStep_ST) - { - ISzAlloc_Free(p->allocMid, p->outBuf); - p->outBufSize = 0; - p->outBuf = (Byte *)ISzAlloc_Alloc(p->allocMid, p->props.outStep_ST); - if (!p->outBuf) - return SZ_ERROR_MEM; - p->outBufSize = p->props.outStep_ST; - } - - if (!p->inBuf || p->inBufSize != p->props.inBufSize_ST) - { - ISzAlloc_Free(p->allocMid, p->inBuf); - p->inBufSize = 0; - p->inBuf = (Byte *)ISzAlloc_Alloc(p->allocMid, p->props.inBufSize_ST); - if (!p->inBuf) - return SZ_ERROR_MEM; - p->inBufSize = p->props.inBufSize_ST; - } - - dec = &p->dec; - dec->decodeToStreamSignature = False; - // dec->decodeOnlyOneBlock = False; - - XzUnpacker_SetOutBuf(dec, NULL, 0); - - inPrev = p->inProcessed; - outPrev = p->outProcessed; - - inPos = 0; - inLim = 0; - inData = NULL; - outPos = 0; - - for (;;) - { - SizeT outSize; - BoolInt finished; - ECoderFinishMode finishMode; - SizeT inProcessed; - ECoderStatus status; - SRes res; - - SizeT outProcessed; - - - - if (inPos == inLim) - { - #ifndef _7ZIP_ST - if (tMode) - { - inData = MtDec_Read(&p->mtc, &inLim); - inPos = 0; - if (inData) - continue; - tMode = False; - inLim = 0; - } - #endif - - if (!p->readWasFinished) - { - inPos = 0; - inLim = p->inBufSize; - inData = p->inBuf; - p->readRes = ISeqInStream_Read(p->inStream, (void *)p->inBuf, &inLim); - p->readProcessed += inLim; - if (inLim == 0 || p->readRes != SZ_OK) - p->readWasFinished = True; - } - } - - outSize = p->props.outStep_ST - outPos; - - finishMode = CODER_FINISH_ANY; - if (p->outSize_Defined) - { - const UInt64 rem = p->outSize - p->outProcessed; - if (outSize >= rem) - { - outSize = (SizeT)rem; - if (p->finishMode) - finishMode = CODER_FINISH_END; - } - } - - inProcessed = inLim - inPos; - outProcessed = outSize; - - res = XzUnpacker_Code(dec, p->outBuf + outPos, &outProcessed, - inData + inPos, &inProcessed, - (inPos == inLim), // srcFinished - finishMode, &status); - - p->codeRes = res; - p->status = status; - - inPos += inProcessed; - outPos += outProcessed; - p->inProcessed += inProcessed; - p->outProcessed += outProcessed; - - finished = ((inProcessed == 0 && outProcessed == 0) || res != SZ_OK); - - if (finished || outProcessed >= outSize) - if (outPos != 0) - { - const size_t written = ISeqOutStream_Write(p->outStream, p->outBuf, outPos); - // p->outProcessed += written; // 21.01: BUG fixed - if (written != outPos) - { - stat->CombinedRes_Type = SZ_ERROR_WRITE; - return SZ_ERROR_WRITE; - } - outPos = 0; - } - - if (p->progress && res == SZ_OK) - { - if (p->inProcessed - inPrev >= (1 << 22) || - p->outProcessed - outPrev >= (1 << 22)) - { - res = ICompressProgress_Progress(p->progress, p->inProcessed, p->outProcessed); - if (res != SZ_OK) - { - stat->CombinedRes_Type = SZ_ERROR_PROGRESS; - stat->ProgressRes = res; - return res; - } - inPrev = p->inProcessed; - outPrev = p->outProcessed; - } - } - - if (finished) - { - // p->codeRes is preliminary error from XzUnpacker_Code. - // and it can be corrected later as final result - // so we return SZ_OK here instead of (res); - return SZ_OK; - // return res; - } - } -} - - - -/* -XzStatInfo_SetStat() transforms - CXzUnpacker return code and status to combined CXzStatInfo results. - it can convert SZ_OK to SZ_ERROR_INPUT_EOF - it can convert SZ_ERROR_NO_ARCHIVE to SZ_OK and (DataAfterEnd = 1) -*/ - -static void XzStatInfo_SetStat(const CXzUnpacker *dec, - int finishMode, - // UInt64 readProcessed, - UInt64 inProcessed, - SRes res, // it's result from CXzUnpacker unpacker - ECoderStatus status, - BoolInt decodingTruncated, - CXzStatInfo *stat) -{ - UInt64 extraSize; - - stat->DecodingTruncated = (Byte)(decodingTruncated ? 1 : 0); - stat->InSize = inProcessed; - stat->NumStreams = dec->numStartedStreams; - stat->NumBlocks = dec->numTotalBlocks; - - stat->UnpackSize_Defined = True; - stat->NumStreams_Defined = True; - stat->NumBlocks_Defined = True; - - extraSize = XzUnpacker_GetExtraSize(dec); - - if (res == SZ_OK) - { - if (status == CODER_STATUS_NEEDS_MORE_INPUT) - { - // CODER_STATUS_NEEDS_MORE_INPUT is expected status for correct xz streams - // any extra data is part of correct data - extraSize = 0; - // if xz stream was not finished, then we need more data - if (!XzUnpacker_IsStreamWasFinished(dec)) - res = SZ_ERROR_INPUT_EOF; - } - else - { - // CODER_STATUS_FINISHED_WITH_MARK is not possible for multi stream xz decoding - // so he we have (status == CODER_STATUS_NOT_FINISHED) - // if (status != CODER_STATUS_FINISHED_WITH_MARK) - if (!decodingTruncated || finishMode) - res = SZ_ERROR_DATA; - } - } - else if (res == SZ_ERROR_NO_ARCHIVE) - { - /* - SZ_ERROR_NO_ARCHIVE is possible for 2 states: - XZ_STATE_STREAM_HEADER - if bad signature or bad CRC - XZ_STATE_STREAM_PADDING - if non-zero padding data - extraSize and inProcessed don't include "bad" byte - */ - // if (inProcessed == extraSize), there was no any good xz stream header, and we keep error - if (inProcessed != extraSize) // if there were good xz streams before error - { - // if (extraSize != 0 || readProcessed != inProcessed) - { - // he we suppose that all xz streams were finsihed OK, and we have - // some extra data after all streams - stat->DataAfterEnd = True; - res = SZ_OK; - } - } - } - - if (stat->DecodeRes == SZ_OK) - stat->DecodeRes = res; - - stat->InSize -= extraSize; -} - - - -SRes XzDecMt_Decode(CXzDecMtHandle pp, - const CXzDecMtProps *props, - const UInt64 *outDataSize, int finishMode, - ISeqOutStream *outStream, - // Byte *outBuf, size_t *outBufSize, - ISeqInStream *inStream, - // const Byte *inData, size_t inDataSize, - CXzStatInfo *stat, - int *isMT, - ICompressProgress *progress) -{ - CXzDecMt *p = (CXzDecMt *)pp; - #ifndef _7ZIP_ST - BoolInt tMode; - #endif - - XzStatInfo_Clear(stat); - - p->props = *props; - - p->inStream = inStream; - p->outStream = outStream; - p->progress = progress; - // p->stat = stat; - - p->outSize = 0; - p->outSize_Defined = False; - if (outDataSize) - { - p->outSize_Defined = True; - p->outSize = *outDataSize; - } - - p->finishMode = finishMode; - - // p->outSize = 457; p->outSize_Defined = True; p->finishMode = False; // for test - - p->writeRes = SZ_OK; - p->outProcessed = 0; - p->inProcessed = 0; - p->readProcessed = 0; - p->readWasFinished = False; - p->readRes = SZ_OK; - - p->codeRes = SZ_OK; - p->status = CODER_STATUS_NOT_SPECIFIED; - - XzUnpacker_Init(&p->dec); - - *isMT = False; - - /* - p->outBuf = NULL; - p->outBufSize = 0; - if (!outStream) - { - p->outBuf = outBuf; - p->outBufSize = *outBufSize; - *outBufSize = 0; - } - */ - - - #ifndef _7ZIP_ST - - p->isBlockHeaderState_Parse = False; - p->isBlockHeaderState_Write = False; - // p->numBadBlocks = 0; - p->mainErrorCode = SZ_OK; - p->mainDecoderWasCalled = False; - - tMode = False; - - if (p->props.numThreads > 1) - { - IMtDecCallback2 vt; - BoolInt needContinue; - SRes res; - // we just free ST buffers here - // but we still keep state variables, that was set in XzUnpacker_Init() - XzDecMt_FreeSt(p); - - p->outProcessed_Parse = 0; - p->parsing_Truncated = False; - - p->numStreams = 0; - p->numTotalBlocks = 0; - p->numBlocks = 0; - p->finishedDecoderIndex = -1; - - if (!p->mtc_WasConstructed) - { - p->mtc_WasConstructed = True; - MtDec_Construct(&p->mtc); - } - - p->mtc.mtCallback = &vt; - p->mtc.mtCallbackObject = p; - - p->mtc.progress = progress; - p->mtc.inStream = inStream; - p->mtc.alloc = &p->alignOffsetAlloc.vt; - // p->mtc.inData = inData; - // p->mtc.inDataSize = inDataSize; - p->mtc.inBufSize = p->props.inBufSize_MT; - // p->mtc.inBlockMax = p->props.inBlockMax; - p->mtc.numThreadsMax = p->props.numThreads; - - *isMT = True; - - vt.Parse = XzDecMt_Callback_Parse; - vt.PreCode = XzDecMt_Callback_PreCode; - vt.Code = XzDecMt_Callback_Code; - vt.Write = XzDecMt_Callback_Write; - - - res = MtDec_Code(&p->mtc); - - - stat->InSize = p->mtc.inProcessed; - - p->inProcessed = p->mtc.inProcessed; - p->readRes = p->mtc.readRes; - p->readWasFinished = p->mtc.readWasFinished; - p->readProcessed = p->mtc.readProcessed; - - tMode = True; - needContinue = False; - - if (res == SZ_OK) - { - if (p->mtc.mtProgress.res != SZ_OK) - { - res = p->mtc.mtProgress.res; - stat->ProgressRes = res; - stat->CombinedRes_Type = SZ_ERROR_PROGRESS; - } - else - needContinue = p->mtc.needContinue; - } - - if (!needContinue) - { - { - SRes codeRes; - BoolInt truncated = False; - ECoderStatus status; - const CXzUnpacker *dec; - - stat->OutSize = p->outProcessed; - - if (p->finishedDecoderIndex >= 0) - { - const CXzDecMtThread *coder = &p->coders[(unsigned)p->finishedDecoderIndex]; - codeRes = coder->codeRes; - dec = &coder->dec; - status = coder->status; - } - else if (p->mainDecoderWasCalled) - { - codeRes = p->codeRes; - dec = &p->dec; - status = p->status; - truncated = p->parsing_Truncated; - } - else - return SZ_ERROR_FAIL; - - if (p->mainErrorCode != SZ_OK) - stat->DecodeRes = p->mainErrorCode; - - XzStatInfo_SetStat(dec, p->finishMode, - // p->mtc.readProcessed, - p->mtc.inProcessed, - codeRes, status, - truncated, - stat); - } - - if (res == SZ_OK) - { - stat->ReadRes = p->mtc.readRes; - - if (p->writeRes != SZ_OK) - { - res = p->writeRes; - stat->CombinedRes_Type = SZ_ERROR_WRITE; - } - else if (p->mtc.readRes != SZ_OK - // && p->mtc.inProcessed == p->mtc.readProcessed - && stat->DecodeRes == SZ_ERROR_INPUT_EOF) - { - res = p->mtc.readRes; - stat->CombinedRes_Type = SZ_ERROR_READ; - } - else if (stat->DecodeRes != SZ_OK) - res = stat->DecodeRes; - } - - stat->CombinedRes = res; - if (stat->CombinedRes_Type == SZ_OK) - stat->CombinedRes_Type = res; - return res; - } - - PRF_STR("----- decoding ST -----"); - } - - #endif - - - *isMT = False; - - { - SRes res = XzDecMt_Decode_ST(p - #ifndef _7ZIP_ST - , tMode - #endif - , stat - ); - - #ifndef _7ZIP_ST - // we must set error code from MT decoding at first - if (p->mainErrorCode != SZ_OK) - stat->DecodeRes = p->mainErrorCode; - #endif - - XzStatInfo_SetStat(&p->dec, - p->finishMode, - // p->readProcessed, - p->inProcessed, - p->codeRes, p->status, - False, // truncated - stat); - - stat->ReadRes = p->readRes; - - if (res == SZ_OK) - { - if (p->readRes != SZ_OK - // && p->inProcessed == p->readProcessed - && stat->DecodeRes == SZ_ERROR_INPUT_EOF) - { - // we set read error as combined error, only if that error was the reason - // of decoding problem - res = p->readRes; - stat->CombinedRes_Type = SZ_ERROR_READ; - } - else if (stat->DecodeRes != SZ_OK) - res = stat->DecodeRes; - } - - stat->CombinedRes = res; - if (stat->CombinedRes_Type == SZ_OK) - stat->CombinedRes_Type = res; - return res; - } -} +/* XzDec.c -- Xz Decode +2021-09-04 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +// #include + +// #define XZ_DUMP + +/* #define XZ_DUMP */ + +#ifdef XZ_DUMP +#include +#endif + +// #define SHOW_DEBUG_INFO + +#ifdef SHOW_DEBUG_INFO +#include +#endif + +#ifdef SHOW_DEBUG_INFO +#define PRF(x) x +#else +#define PRF(x) +#endif + +#define PRF_STR(s) PRF(printf("\n" s "\n")) +#define PRF_STR_INT(s, d) PRF(printf("\n" s " %d\n", (unsigned)d)) + +#include +#include + +#include "7zCrc.h" +#include "Alloc.h" +#include "Bra.h" +#include "CpuArch.h" +#include "Delta.h" +#include "Lzma2Dec.h" + +// #define USE_SUBBLOCK + +#ifdef USE_SUBBLOCK +#include "Bcj3Dec.c" +#include "SbDec.h" +#endif + +#include "Xz.h" + +#define XZ_CHECK_SIZE_MAX 64 + +#define CODER_BUF_SIZE ((size_t)1 << 17) + +unsigned Xz_ReadVarInt(const Byte *p, size_t maxSize, UInt64 *value) +{ + unsigned i, limit; + *value = 0; + limit = (maxSize > 9) ? 9 : (unsigned)maxSize; + + for (i = 0; i < limit;) + { + Byte b = p[i]; + *value |= (UInt64)(b & 0x7F) << (7 * i++); + if ((b & 0x80) == 0) + return (b == 0 && i != 1) ? 0 : i; + } + return 0; +} + +/* ---------- BraState ---------- */ + +#define BRA_BUF_SIZE (1 << 14) + +typedef struct +{ + size_t bufPos; + size_t bufConv; + size_t bufTotal; + + int encodeMode; + + UInt32 methodId; + UInt32 delta; + UInt32 ip; + UInt32 x86State; + Byte deltaState[DELTA_STATE_SIZE]; + + Byte buf[BRA_BUF_SIZE]; +} CBraState; + +static void BraState_Free(void *pp, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, pp); +} + +static SRes BraState_SetProps(void *pp, const Byte *props, size_t propSize, ISzAllocPtr alloc) +{ + CBraState *p = ((CBraState *)pp); + UNUSED_VAR(alloc); + p->ip = 0; + if (p->methodId == XZ_ID_Delta) + { + if (propSize != 1) + return SZ_ERROR_UNSUPPORTED; + p->delta = (unsigned)props[0] + 1; + } + else + { + if (propSize == 4) + { + UInt32 v = GetUi32(props); + switch (p->methodId) + { + case XZ_ID_PPC: + case XZ_ID_ARM: + case XZ_ID_SPARC: + if ((v & 3) != 0) + return SZ_ERROR_UNSUPPORTED; + break; + case XZ_ID_ARMT: + if ((v & 1) != 0) + return SZ_ERROR_UNSUPPORTED; + break; + case XZ_ID_IA64: + if ((v & 0xF) != 0) + return SZ_ERROR_UNSUPPORTED; + break; + } + p->ip = v; + } + else if (propSize != 0) + return SZ_ERROR_UNSUPPORTED; + } + return SZ_OK; +} + +static void BraState_Init(void *pp) +{ + CBraState *p = ((CBraState *)pp); + p->bufPos = p->bufConv = p->bufTotal = 0; + x86_Convert_Init(p->x86State); + if (p->methodId == XZ_ID_Delta) + Delta_Init(p->deltaState); +} + + +#define CASE_BRA_CONV(isa) case XZ_ID_ ## isa: size = isa ## _Convert(data, size, p->ip, p->encodeMode); break; + +static SizeT BraState_Filter(void *pp, Byte *data, SizeT size) +{ + CBraState *p = ((CBraState *)pp); + switch (p->methodId) + { + case XZ_ID_Delta: + if (p->encodeMode) + Delta_Encode(p->deltaState, p->delta, data, size); + else + Delta_Decode(p->deltaState, p->delta, data, size); + break; + case XZ_ID_X86: + size = x86_Convert(data, size, p->ip, &p->x86State, p->encodeMode); + break; + CASE_BRA_CONV(PPC) + CASE_BRA_CONV(IA64) + CASE_BRA_CONV(ARM) + CASE_BRA_CONV(ARMT) + CASE_BRA_CONV(SPARC) + } + p->ip += (UInt32)size; + return size; +} + + +static SRes BraState_Code2(void *pp, + Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, int srcWasFinished, + ECoderFinishMode finishMode, + // int *wasFinished + ECoderStatus *status) +{ + CBraState *p = ((CBraState *)pp); + SizeT destRem = *destLen; + SizeT srcRem = *srcLen; + UNUSED_VAR(finishMode); + + *destLen = 0; + *srcLen = 0; + // *wasFinished = False; + *status = CODER_STATUS_NOT_FINISHED; + + while (destRem > 0) + { + if (p->bufPos != p->bufConv) + { + size_t size = p->bufConv - p->bufPos; + if (size > destRem) + size = destRem; + memcpy(dest, p->buf + p->bufPos, size); + p->bufPos += size; + *destLen += size; + dest += size; + destRem -= size; + continue; + } + + p->bufTotal -= p->bufPos; + memmove(p->buf, p->buf + p->bufPos, p->bufTotal); + p->bufPos = 0; + p->bufConv = 0; + { + size_t size = BRA_BUF_SIZE - p->bufTotal; + if (size > srcRem) + size = srcRem; + memcpy(p->buf + p->bufTotal, src, size); + *srcLen += size; + src += size; + srcRem -= size; + p->bufTotal += size; + } + if (p->bufTotal == 0) + break; + + p->bufConv = BraState_Filter(pp, p->buf, p->bufTotal); + + if (p->bufConv == 0) + { + if (!srcWasFinished) + break; + p->bufConv = p->bufTotal; + } + } + + if (p->bufTotal == p->bufPos && srcRem == 0 && srcWasFinished) + { + *status = CODER_STATUS_FINISHED_WITH_MARK; + // *wasFinished = 1; + } + + return SZ_OK; +} + + +SRes BraState_SetFromMethod(IStateCoder *p, UInt64 id, int encodeMode, ISzAllocPtr alloc); +SRes BraState_SetFromMethod(IStateCoder *p, UInt64 id, int encodeMode, ISzAllocPtr alloc) +{ + CBraState *decoder; + if (id < XZ_ID_Delta || id > XZ_ID_SPARC) + return SZ_ERROR_UNSUPPORTED; + decoder = (CBraState *)p->p; + if (!decoder) + { + decoder = (CBraState *)ISzAlloc_Alloc(alloc, sizeof(CBraState)); + if (!decoder) + return SZ_ERROR_MEM; + p->p = decoder; + p->Free = BraState_Free; + p->SetProps = BraState_SetProps; + p->Init = BraState_Init; + p->Code2 = BraState_Code2; + p->Filter = BraState_Filter; + } + decoder->methodId = (UInt32)id; + decoder->encodeMode = encodeMode; + return SZ_OK; +} + + + +/* ---------- SbState ---------- */ + +#ifdef USE_SUBBLOCK + +static void SbState_Free(void *pp, ISzAllocPtr alloc) +{ + CSbDec *p = (CSbDec *)pp; + SbDec_Free(p); + ISzAlloc_Free(alloc, pp); +} + +static SRes SbState_SetProps(void *pp, const Byte *props, size_t propSize, ISzAllocPtr alloc) +{ + UNUSED_VAR(pp); + UNUSED_VAR(props); + UNUSED_VAR(alloc); + return (propSize == 0) ? SZ_OK : SZ_ERROR_UNSUPPORTED; +} + +static void SbState_Init(void *pp) +{ + SbDec_Init((CSbDec *)pp); +} + +static SRes SbState_Code2(void *pp, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + int srcWasFinished, ECoderFinishMode finishMode, + // int *wasFinished + ECoderStatus *status) +{ + CSbDec *p = (CSbDec *)pp; + SRes res; + UNUSED_VAR(srcWasFinished); + p->dest = dest; + p->destLen = *destLen; + p->src = src; + p->srcLen = *srcLen; + p->finish = finishMode; /* change it */ + res = SbDec_Decode((CSbDec *)pp); + *destLen -= p->destLen; + *srcLen -= p->srcLen; + // *wasFinished = (*destLen == 0 && *srcLen == 0); /* change it */ + *status = (*destLen == 0 && *srcLen == 0) ? + CODER_STATUS_FINISHED_WITH_MARK : + CODER_STATUS_NOT_FINISHED; + return res; +} + +static SRes SbState_SetFromMethod(IStateCoder *p, ISzAllocPtr alloc) +{ + CSbDec *decoder = (CSbDec *)p->p; + if (!decoder) + { + decoder = (CSbDec *)ISzAlloc_Alloc(alloc, sizeof(CSbDec)); + if (!decoder) + return SZ_ERROR_MEM; + p->p = decoder; + p->Free = SbState_Free; + p->SetProps = SbState_SetProps; + p->Init = SbState_Init; + p->Code2 = SbState_Code2; + p->Filter = NULL; + } + SbDec_Construct(decoder); + SbDec_SetAlloc(decoder, alloc); + return SZ_OK; +} + +#endif + + + +/* ---------- Lzma2 ---------- */ + +typedef struct +{ + CLzma2Dec decoder; + BoolInt outBufMode; +} CLzma2Dec_Spec; + + +static void Lzma2State_Free(void *pp, ISzAllocPtr alloc) +{ + CLzma2Dec_Spec *p = (CLzma2Dec_Spec *)pp; + if (p->outBufMode) + Lzma2Dec_FreeProbs(&p->decoder, alloc); + else + Lzma2Dec_Free(&p->decoder, alloc); + ISzAlloc_Free(alloc, pp); +} + +static SRes Lzma2State_SetProps(void *pp, const Byte *props, size_t propSize, ISzAllocPtr alloc) +{ + if (propSize != 1) + return SZ_ERROR_UNSUPPORTED; + { + CLzma2Dec_Spec *p = (CLzma2Dec_Spec *)pp; + if (p->outBufMode) + return Lzma2Dec_AllocateProbs(&p->decoder, props[0], alloc); + else + return Lzma2Dec_Allocate(&p->decoder, props[0], alloc); + } +} + +static void Lzma2State_Init(void *pp) +{ + Lzma2Dec_Init(&((CLzma2Dec_Spec *)pp)->decoder); +} + + +/* + if (outBufMode), then (dest) is not used. Use NULL. + Data is unpacked to (spec->decoder.decoder.dic) output buffer. +*/ + +static SRes Lzma2State_Code2(void *pp, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + int srcWasFinished, ECoderFinishMode finishMode, + // int *wasFinished, + ECoderStatus *status) +{ + CLzma2Dec_Spec *spec = (CLzma2Dec_Spec *)pp; + ELzmaStatus status2; + /* ELzmaFinishMode fm = (finishMode == LZMA_FINISH_ANY) ? LZMA_FINISH_ANY : LZMA_FINISH_END; */ + SRes res; + UNUSED_VAR(srcWasFinished); + if (spec->outBufMode) + { + SizeT dicPos = spec->decoder.decoder.dicPos; + SizeT dicLimit = dicPos + *destLen; + res = Lzma2Dec_DecodeToDic(&spec->decoder, dicLimit, src, srcLen, (ELzmaFinishMode)finishMode, &status2); + *destLen = spec->decoder.decoder.dicPos - dicPos; + } + else + res = Lzma2Dec_DecodeToBuf(&spec->decoder, dest, destLen, src, srcLen, (ELzmaFinishMode)finishMode, &status2); + // *wasFinished = (status2 == LZMA_STATUS_FINISHED_WITH_MARK); + // ECoderStatus values are identical to ELzmaStatus values of LZMA2 decoder + *status = (ECoderStatus)status2; + return res; +} + + +static SRes Lzma2State_SetFromMethod(IStateCoder *p, Byte *outBuf, size_t outBufSize, ISzAllocPtr alloc) +{ + CLzma2Dec_Spec *spec = (CLzma2Dec_Spec *)p->p; + if (!spec) + { + spec = (CLzma2Dec_Spec *)ISzAlloc_Alloc(alloc, sizeof(CLzma2Dec_Spec)); + if (!spec) + return SZ_ERROR_MEM; + p->p = spec; + p->Free = Lzma2State_Free; + p->SetProps = Lzma2State_SetProps; + p->Init = Lzma2State_Init; + p->Code2 = Lzma2State_Code2; + p->Filter = NULL; + Lzma2Dec_Construct(&spec->decoder); + } + spec->outBufMode = False; + if (outBuf) + { + spec->outBufMode = True; + spec->decoder.decoder.dic = outBuf; + spec->decoder.decoder.dicBufSize = outBufSize; + } + return SZ_OK; +} + + +static SRes Lzma2State_ResetOutBuf(IStateCoder *p, Byte *outBuf, size_t outBufSize) +{ + CLzma2Dec_Spec *spec = (CLzma2Dec_Spec *)p->p; + if ((spec->outBufMode && !outBuf) || (!spec->outBufMode && outBuf)) + return SZ_ERROR_FAIL; + if (outBuf) + { + spec->decoder.decoder.dic = outBuf; + spec->decoder.decoder.dicBufSize = outBufSize; + } + return SZ_OK; +} + + + +static void MixCoder_Construct(CMixCoder *p, ISzAllocPtr alloc) +{ + unsigned i; + p->alloc = alloc; + p->buf = NULL; + p->numCoders = 0; + + p->outBufSize = 0; + p->outBuf = NULL; + // p->SingleBufMode = False; + + for (i = 0; i < MIXCODER_NUM_FILTERS_MAX; i++) + p->coders[i].p = NULL; +} + + +static void MixCoder_Free(CMixCoder *p) +{ + unsigned i; + p->numCoders = 0; + for (i = 0; i < MIXCODER_NUM_FILTERS_MAX; i++) + { + IStateCoder *sc = &p->coders[i]; + if (sc->p) + { + sc->Free(sc->p, p->alloc); + sc->p = NULL; + } + } + if (p->buf) + { + ISzAlloc_Free(p->alloc, p->buf); + p->buf = NULL; /* 9.31: the BUG was fixed */ + } +} + +static void MixCoder_Init(CMixCoder *p) +{ + unsigned i; + for (i = 0; i < MIXCODER_NUM_FILTERS_MAX - 1; i++) + { + p->size[i] = 0; + p->pos[i] = 0; + p->finished[i] = 0; + } + for (i = 0; i < p->numCoders; i++) + { + IStateCoder *coder = &p->coders[i]; + coder->Init(coder->p); + p->results[i] = SZ_OK; + } + p->outWritten = 0; + p->wasFinished = False; + p->res = SZ_OK; + p->status = CODER_STATUS_NOT_SPECIFIED; +} + + +static SRes MixCoder_SetFromMethod(CMixCoder *p, unsigned coderIndex, UInt64 methodId, Byte *outBuf, size_t outBufSize) +{ + IStateCoder *sc = &p->coders[coderIndex]; + p->ids[coderIndex] = methodId; + switch (methodId) + { + case XZ_ID_LZMA2: return Lzma2State_SetFromMethod(sc, outBuf, outBufSize, p->alloc); + #ifdef USE_SUBBLOCK + case XZ_ID_Subblock: return SbState_SetFromMethod(sc, p->alloc); + #endif + } + if (coderIndex == 0) + return SZ_ERROR_UNSUPPORTED; + return BraState_SetFromMethod(sc, methodId, 0, p->alloc); +} + + +static SRes MixCoder_ResetFromMethod(CMixCoder *p, unsigned coderIndex, UInt64 methodId, Byte *outBuf, size_t outBufSize) +{ + IStateCoder *sc = &p->coders[coderIndex]; + switch (methodId) + { + case XZ_ID_LZMA2: return Lzma2State_ResetOutBuf(sc, outBuf, outBufSize); + } + return SZ_ERROR_UNSUPPORTED; +} + + + +/* + if (destFinish) - then unpack data block is finished at (*destLen) position, + and we can return data that were not processed by filter + +output (status) can be : + CODER_STATUS_NOT_FINISHED + CODER_STATUS_FINISHED_WITH_MARK + CODER_STATUS_NEEDS_MORE_INPUT - not implemented still +*/ + +static SRes MixCoder_Code(CMixCoder *p, + Byte *dest, SizeT *destLen, int destFinish, + const Byte *src, SizeT *srcLen, int srcWasFinished, + ECoderFinishMode finishMode) +{ + SizeT destLenOrig = *destLen; + SizeT srcLenOrig = *srcLen; + + *destLen = 0; + *srcLen = 0; + + if (p->wasFinished) + return p->res; + + p->status = CODER_STATUS_NOT_FINISHED; + + // if (p->SingleBufMode) + if (p->outBuf) + { + SRes res; + SizeT destLen2, srcLen2; + int wasFinished; + + PRF_STR("------- MixCoder Single ----------"); + + srcLen2 = srcLenOrig; + destLen2 = destLenOrig; + + { + IStateCoder *coder = &p->coders[0]; + res = coder->Code2(coder->p, NULL, &destLen2, src, &srcLen2, srcWasFinished, finishMode, + // &wasFinished, + &p->status); + wasFinished = (p->status == CODER_STATUS_FINISHED_WITH_MARK); + } + + p->res = res; + + /* + if (wasFinished) + p->status = CODER_STATUS_FINISHED_WITH_MARK; + else + { + if (res == SZ_OK) + if (destLen2 != destLenOrig) + p->status = CODER_STATUS_NEEDS_MORE_INPUT; + } + */ + + + *srcLen = srcLen2; + src += srcLen2; + p->outWritten += destLen2; + + if (res != SZ_OK || srcWasFinished || wasFinished) + p->wasFinished = True; + + if (p->numCoders == 1) + *destLen = destLen2; + else if (p->wasFinished) + { + unsigned i; + size_t processed = p->outWritten; + + for (i = 1; i < p->numCoders; i++) + { + IStateCoder *coder = &p->coders[i]; + processed = coder->Filter(coder->p, p->outBuf, processed); + if (wasFinished || (destFinish && p->outWritten == destLenOrig)) + processed = p->outWritten; + PRF_STR_INT("filter", i); + } + *destLen = processed; + } + return res; + } + + PRF_STR("standard mix"); + + if (p->numCoders != 1) + { + if (!p->buf) + { + p->buf = (Byte *)ISzAlloc_Alloc(p->alloc, CODER_BUF_SIZE * (MIXCODER_NUM_FILTERS_MAX - 1)); + if (!p->buf) + return SZ_ERROR_MEM; + } + + finishMode = CODER_FINISH_ANY; + } + + for (;;) + { + BoolInt processed = False; + BoolInt allFinished = True; + SRes resMain = SZ_OK; + unsigned i; + + p->status = CODER_STATUS_NOT_FINISHED; + /* + if (p->numCoders == 1 && *destLen == destLenOrig && finishMode == LZMA_FINISH_ANY) + break; + */ + + for (i = 0; i < p->numCoders; i++) + { + SRes res; + IStateCoder *coder = &p->coders[i]; + Byte *dest2; + SizeT destLen2, srcLen2; // destLen2_Orig; + const Byte *src2; + int srcFinished2; + int encodingWasFinished; + ECoderStatus status2; + + if (i == 0) + { + src2 = src; + srcLen2 = srcLenOrig - *srcLen; + srcFinished2 = srcWasFinished; + } + else + { + size_t k = i - 1; + src2 = p->buf + (CODER_BUF_SIZE * k) + p->pos[k]; + srcLen2 = p->size[k] - p->pos[k]; + srcFinished2 = p->finished[k]; + } + + if (i == p->numCoders - 1) + { + dest2 = dest; + destLen2 = destLenOrig - *destLen; + } + else + { + if (p->pos[i] != p->size[i]) + continue; + dest2 = p->buf + (CODER_BUF_SIZE * i); + destLen2 = CODER_BUF_SIZE; + } + + // destLen2_Orig = destLen2; + + if (p->results[i] != SZ_OK) + { + if (resMain == SZ_OK) + resMain = p->results[i]; + continue; + } + + res = coder->Code2(coder->p, + dest2, &destLen2, + src2, &srcLen2, srcFinished2, + finishMode, + // &encodingWasFinished, + &status2); + + if (res != SZ_OK) + { + p->results[i] = res; + if (resMain == SZ_OK) + resMain = res; + } + + encodingWasFinished = (status2 == CODER_STATUS_FINISHED_WITH_MARK); + + if (!encodingWasFinished) + { + allFinished = False; + if (p->numCoders == 1 && res == SZ_OK) + p->status = status2; + } + + if (i == 0) + { + *srcLen += srcLen2; + src += srcLen2; + } + else + p->pos[(size_t)i - 1] += srcLen2; + + if (i == p->numCoders - 1) + { + *destLen += destLen2; + dest += destLen2; + } + else + { + p->size[i] = destLen2; + p->pos[i] = 0; + p->finished[i] = encodingWasFinished; + } + + if (destLen2 != 0 || srcLen2 != 0) + processed = True; + } + + if (!processed) + { + if (allFinished) + p->status = CODER_STATUS_FINISHED_WITH_MARK; + return resMain; + } + } +} + + +SRes Xz_ParseHeader(CXzStreamFlags *p, const Byte *buf) +{ + *p = (CXzStreamFlags)GetBe16(buf + XZ_SIG_SIZE); + if (CrcCalc(buf + XZ_SIG_SIZE, XZ_STREAM_FLAGS_SIZE) != + GetUi32(buf + XZ_SIG_SIZE + XZ_STREAM_FLAGS_SIZE)) + return SZ_ERROR_NO_ARCHIVE; + return XzFlags_IsSupported(*p) ? SZ_OK : SZ_ERROR_UNSUPPORTED; +} + +static BoolInt Xz_CheckFooter(CXzStreamFlags flags, UInt64 indexSize, const Byte *buf) +{ + return indexSize == (((UInt64)GetUi32(buf + 4) + 1) << 2) + && GetUi32(buf) == CrcCalc(buf + 4, 6) + && flags == GetBe16(buf + 8) + && buf[10] == XZ_FOOTER_SIG_0 + && buf[11] == XZ_FOOTER_SIG_1; +} + +#define READ_VARINT_AND_CHECK(buf, pos, size, res) \ + { unsigned s = Xz_ReadVarInt(buf + pos, size - pos, res); \ + if (s == 0) return SZ_ERROR_ARCHIVE; \ + pos += s; } + + +static BoolInt XzBlock_AreSupportedFilters(const CXzBlock *p) +{ + unsigned numFilters = XzBlock_GetNumFilters(p) - 1; + unsigned i; + { + const CXzFilter *f = &p->filters[numFilters]; + if (f->id != XZ_ID_LZMA2 || f->propsSize != 1 || f->props[0] > 40) + return False; + } + + for (i = 0; i < numFilters; i++) + { + const CXzFilter *f = &p->filters[i]; + if (f->id == XZ_ID_Delta) + { + if (f->propsSize != 1) + return False; + } + else if (f->id < XZ_ID_Delta + || f->id > XZ_ID_SPARC + || (f->propsSize != 0 && f->propsSize != 4)) + return False; + } + return True; +} + + +SRes XzBlock_Parse(CXzBlock *p, const Byte *header) +{ + unsigned pos; + unsigned numFilters, i; + unsigned headerSize = (unsigned)header[0] << 2; + + /* (headerSize != 0) : another code checks */ + + if (CrcCalc(header, headerSize) != GetUi32(header + headerSize)) + return SZ_ERROR_ARCHIVE; + + pos = 1; + p->flags = header[pos++]; + + p->packSize = (UInt64)(Int64)-1; + if (XzBlock_HasPackSize(p)) + { + READ_VARINT_AND_CHECK(header, pos, headerSize, &p->packSize); + if (p->packSize == 0 || p->packSize + headerSize >= (UInt64)1 << 63) + return SZ_ERROR_ARCHIVE; + } + + p->unpackSize = (UInt64)(Int64)-1; + if (XzBlock_HasUnpackSize(p)) + READ_VARINT_AND_CHECK(header, pos, headerSize, &p->unpackSize); + + numFilters = XzBlock_GetNumFilters(p); + for (i = 0; i < numFilters; i++) + { + CXzFilter *filter = p->filters + i; + UInt64 size; + READ_VARINT_AND_CHECK(header, pos, headerSize, &filter->id); + READ_VARINT_AND_CHECK(header, pos, headerSize, &size); + if (size > headerSize - pos || size > XZ_FILTER_PROPS_SIZE_MAX) + return SZ_ERROR_ARCHIVE; + filter->propsSize = (UInt32)size; + memcpy(filter->props, header + pos, (size_t)size); + pos += (unsigned)size; + + #ifdef XZ_DUMP + printf("\nf[%u] = %2X: ", i, (unsigned)filter->id); + { + unsigned i; + for (i = 0; i < size; i++) + printf(" %2X", filter->props[i]); + } + #endif + } + + if (XzBlock_HasUnsupportedFlags(p)) + return SZ_ERROR_UNSUPPORTED; + + while (pos < headerSize) + if (header[pos++] != 0) + return SZ_ERROR_ARCHIVE; + return SZ_OK; +} + + + + +static SRes XzDecMix_Init(CMixCoder *p, const CXzBlock *block, Byte *outBuf, size_t outBufSize) +{ + unsigned i; + BoolInt needReInit = True; + unsigned numFilters = XzBlock_GetNumFilters(block); + + if (numFilters == p->numCoders && ((p->outBuf && outBuf) || (!p->outBuf && !outBuf))) + { + needReInit = False; + for (i = 0; i < numFilters; i++) + if (p->ids[i] != block->filters[numFilters - 1 - i].id) + { + needReInit = True; + break; + } + } + + // p->SingleBufMode = (outBuf != NULL); + p->outBuf = outBuf; + p->outBufSize = outBufSize; + + // p->SingleBufMode = False; + // outBuf = NULL; + + if (needReInit) + { + MixCoder_Free(p); + for (i = 0; i < numFilters; i++) + { + RINOK(MixCoder_SetFromMethod(p, i, block->filters[numFilters - 1 - i].id, outBuf, outBufSize)); + } + p->numCoders = numFilters; + } + else + { + RINOK(MixCoder_ResetFromMethod(p, 0, block->filters[numFilters - 1].id, outBuf, outBufSize)); + } + + for (i = 0; i < numFilters; i++) + { + const CXzFilter *f = &block->filters[numFilters - 1 - i]; + IStateCoder *sc = &p->coders[i]; + RINOK(sc->SetProps(sc->p, f->props, f->propsSize, p->alloc)); + } + + MixCoder_Init(p); + return SZ_OK; +} + + + +void XzUnpacker_Init(CXzUnpacker *p) +{ + p->state = XZ_STATE_STREAM_HEADER; + p->pos = 0; + p->numStartedStreams = 0; + p->numFinishedStreams = 0; + p->numTotalBlocks = 0; + p->padSize = 0; + p->decodeOnlyOneBlock = 0; + + p->parseMode = False; + p->decodeToStreamSignature = False; + + // p->outBuf = NULL; + // p->outBufSize = 0; + p->outDataWritten = 0; +} + + +void XzUnpacker_SetOutBuf(CXzUnpacker *p, Byte *outBuf, size_t outBufSize) +{ + p->outBuf = outBuf; + p->outBufSize = outBufSize; +} + + +void XzUnpacker_Construct(CXzUnpacker *p, ISzAllocPtr alloc) +{ + MixCoder_Construct(&p->decoder, alloc); + p->outBuf = NULL; + p->outBufSize = 0; + XzUnpacker_Init(p); +} + + +void XzUnpacker_Free(CXzUnpacker *p) +{ + MixCoder_Free(&p->decoder); +} + + +void XzUnpacker_PrepareToRandomBlockDecoding(CXzUnpacker *p) +{ + p->indexSize = 0; + p->numBlocks = 0; + Sha256_Init(&p->sha); + p->state = XZ_STATE_BLOCK_HEADER; + p->pos = 0; + p->decodeOnlyOneBlock = 1; +} + + +static void XzUnpacker_UpdateIndex(CXzUnpacker *p, UInt64 packSize, UInt64 unpackSize) +{ + Byte temp[32]; + unsigned num = Xz_WriteVarInt(temp, packSize); + num += Xz_WriteVarInt(temp + num, unpackSize); + Sha256_Update(&p->sha, temp, num); + p->indexSize += num; + p->numBlocks++; +} + + + +SRes XzUnpacker_Code(CXzUnpacker *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, int srcFinished, + ECoderFinishMode finishMode, ECoderStatus *status) +{ + SizeT destLenOrig = *destLen; + SizeT srcLenOrig = *srcLen; + *destLen = 0; + *srcLen = 0; + *status = CODER_STATUS_NOT_SPECIFIED; + + for (;;) + { + SizeT srcRem; + + if (p->state == XZ_STATE_BLOCK) + { + SizeT destLen2 = destLenOrig - *destLen; + SizeT srcLen2 = srcLenOrig - *srcLen; + SRes res; + + ECoderFinishMode finishMode2 = finishMode; + BoolInt srcFinished2 = srcFinished; + BoolInt destFinish = False; + + if (p->block.packSize != (UInt64)(Int64)-1) + { + UInt64 rem = p->block.packSize - p->packSize; + if (srcLen2 >= rem) + { + srcFinished2 = True; + srcLen2 = (SizeT)rem; + } + if (rem == 0 && p->block.unpackSize == p->unpackSize) + return SZ_ERROR_DATA; + } + + if (p->block.unpackSize != (UInt64)(Int64)-1) + { + UInt64 rem = p->block.unpackSize - p->unpackSize; + if (destLen2 >= rem) + { + destFinish = True; + finishMode2 = CODER_FINISH_END; + destLen2 = (SizeT)rem; + } + } + + /* + if (srcLen2 == 0 && destLen2 == 0) + { + *status = CODER_STATUS_NOT_FINISHED; + return SZ_OK; + } + */ + + { + res = MixCoder_Code(&p->decoder, + (p->outBuf ? NULL : dest), &destLen2, destFinish, + src, &srcLen2, srcFinished2, + finishMode2); + + *status = p->decoder.status; + XzCheck_Update(&p->check, (p->outBuf ? p->outBuf + p->outDataWritten : dest), destLen2); + if (!p->outBuf) + dest += destLen2; + p->outDataWritten += destLen2; + } + + (*srcLen) += srcLen2; + src += srcLen2; + p->packSize += srcLen2; + (*destLen) += destLen2; + p->unpackSize += destLen2; + + RINOK(res); + + if (*status != CODER_STATUS_FINISHED_WITH_MARK) + { + if (p->block.packSize == p->packSize + && *status == CODER_STATUS_NEEDS_MORE_INPUT) + { + PRF_STR("CODER_STATUS_NEEDS_MORE_INPUT"); + *status = CODER_STATUS_NOT_SPECIFIED; + return SZ_ERROR_DATA; + } + + return SZ_OK; + } + { + XzUnpacker_UpdateIndex(p, XzUnpacker_GetPackSizeForIndex(p), p->unpackSize); + p->state = XZ_STATE_BLOCK_FOOTER; + p->pos = 0; + p->alignPos = 0; + *status = CODER_STATUS_NOT_SPECIFIED; + + if ((p->block.packSize != (UInt64)(Int64)-1 && p->block.packSize != p->packSize) + || (p->block.unpackSize != (UInt64)(Int64)-1 && p->block.unpackSize != p->unpackSize)) + { + PRF_STR("ERROR: block.size mismatch"); + return SZ_ERROR_DATA; + } + } + // continue; + } + + srcRem = srcLenOrig - *srcLen; + + // XZ_STATE_BLOCK_FOOTER can transit to XZ_STATE_BLOCK_HEADER without input bytes + if (srcRem == 0 && p->state != XZ_STATE_BLOCK_FOOTER) + { + *status = CODER_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + + switch (p->state) + { + case XZ_STATE_STREAM_HEADER: + { + if (p->pos < XZ_STREAM_HEADER_SIZE) + { + if (p->pos < XZ_SIG_SIZE && *src != XZ_SIG[p->pos]) + return SZ_ERROR_NO_ARCHIVE; + if (p->decodeToStreamSignature) + return SZ_OK; + p->buf[p->pos++] = *src++; + (*srcLen)++; + } + else + { + RINOK(Xz_ParseHeader(&p->streamFlags, p->buf)); + p->numStartedStreams++; + p->indexSize = 0; + p->numBlocks = 0; + Sha256_Init(&p->sha); + p->state = XZ_STATE_BLOCK_HEADER; + p->pos = 0; + } + break; + } + + case XZ_STATE_BLOCK_HEADER: + { + if (p->pos == 0) + { + p->buf[p->pos++] = *src++; + (*srcLen)++; + if (p->buf[0] == 0) + { + if (p->decodeOnlyOneBlock) + return SZ_ERROR_DATA; + p->indexPreSize = 1 + Xz_WriteVarInt(p->buf + 1, p->numBlocks); + p->indexPos = p->indexPreSize; + p->indexSize += p->indexPreSize; + Sha256_Final(&p->sha, p->shaDigest); + Sha256_Init(&p->sha); + p->crc = CrcUpdate(CRC_INIT_VAL, p->buf, p->indexPreSize); + p->state = XZ_STATE_STREAM_INDEX; + break; + } + p->blockHeaderSize = ((UInt32)p->buf[0] << 2) + 4; + break; + } + + if (p->pos != p->blockHeaderSize) + { + UInt32 cur = p->blockHeaderSize - p->pos; + if (cur > srcRem) + cur = (UInt32)srcRem; + memcpy(p->buf + p->pos, src, cur); + p->pos += cur; + (*srcLen) += cur; + src += cur; + } + else + { + RINOK(XzBlock_Parse(&p->block, p->buf)); + if (!XzBlock_AreSupportedFilters(&p->block)) + return SZ_ERROR_UNSUPPORTED; + p->numTotalBlocks++; + p->state = XZ_STATE_BLOCK; + p->packSize = 0; + p->unpackSize = 0; + XzCheck_Init(&p->check, XzFlags_GetCheckType(p->streamFlags)); + if (p->parseMode) + { + p->headerParsedOk = True; + return SZ_OK; + } + RINOK(XzDecMix_Init(&p->decoder, &p->block, p->outBuf, p->outBufSize)); + } + break; + } + + case XZ_STATE_BLOCK_FOOTER: + { + if ((((unsigned)p->packSize + p->alignPos) & 3) != 0) + { + if (srcRem == 0) + { + *status = CODER_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + (*srcLen)++; + p->alignPos++; + if (*src++ != 0) + return SZ_ERROR_CRC; + } + else + { + UInt32 checkSize = XzFlags_GetCheckSize(p->streamFlags); + UInt32 cur = checkSize - p->pos; + if (cur != 0) + { + if (srcRem == 0) + { + *status = CODER_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + if (cur > srcRem) + cur = (UInt32)srcRem; + memcpy(p->buf + p->pos, src, cur); + p->pos += cur; + (*srcLen) += cur; + src += cur; + if (checkSize != p->pos) + break; + } + { + Byte digest[XZ_CHECK_SIZE_MAX]; + p->state = XZ_STATE_BLOCK_HEADER; + p->pos = 0; + if (XzCheck_Final(&p->check, digest) && memcmp(digest, p->buf, checkSize) != 0) + return SZ_ERROR_CRC; + if (p->decodeOnlyOneBlock) + { + *status = CODER_STATUS_FINISHED_WITH_MARK; + return SZ_OK; + } + } + } + break; + } + + case XZ_STATE_STREAM_INDEX: + { + if (p->pos < p->indexPreSize) + { + (*srcLen)++; + if (*src++ != p->buf[p->pos++]) + return SZ_ERROR_CRC; + } + else + { + if (p->indexPos < p->indexSize) + { + UInt64 cur = p->indexSize - p->indexPos; + if (srcRem > cur) + srcRem = (SizeT)cur; + p->crc = CrcUpdate(p->crc, src, srcRem); + Sha256_Update(&p->sha, src, srcRem); + (*srcLen) += srcRem; + src += srcRem; + p->indexPos += srcRem; + } + else if ((p->indexPos & 3) != 0) + { + Byte b = *src++; + p->crc = CRC_UPDATE_BYTE(p->crc, b); + (*srcLen)++; + p->indexPos++; + p->indexSize++; + if (b != 0) + return SZ_ERROR_CRC; + } + else + { + Byte digest[SHA256_DIGEST_SIZE]; + p->state = XZ_STATE_STREAM_INDEX_CRC; + p->indexSize += 4; + p->pos = 0; + Sha256_Final(&p->sha, digest); + if (memcmp(digest, p->shaDigest, SHA256_DIGEST_SIZE) != 0) + return SZ_ERROR_CRC; + } + } + break; + } + + case XZ_STATE_STREAM_INDEX_CRC: + { + if (p->pos < 4) + { + (*srcLen)++; + p->buf[p->pos++] = *src++; + } + else + { + const Byte *ptr = p->buf; + p->state = XZ_STATE_STREAM_FOOTER; + p->pos = 0; + if (CRC_GET_DIGEST(p->crc) != GetUi32(ptr)) + return SZ_ERROR_CRC; + } + break; + } + + case XZ_STATE_STREAM_FOOTER: + { + UInt32 cur = XZ_STREAM_FOOTER_SIZE - p->pos; + if (cur > srcRem) + cur = (UInt32)srcRem; + memcpy(p->buf + p->pos, src, cur); + p->pos += cur; + (*srcLen) += cur; + src += cur; + if (p->pos == XZ_STREAM_FOOTER_SIZE) + { + p->state = XZ_STATE_STREAM_PADDING; + p->numFinishedStreams++; + p->padSize = 0; + if (!Xz_CheckFooter(p->streamFlags, p->indexSize, p->buf)) + return SZ_ERROR_CRC; + } + break; + } + + case XZ_STATE_STREAM_PADDING: + { + if (*src != 0) + { + if (((UInt32)p->padSize & 3) != 0) + return SZ_ERROR_NO_ARCHIVE; + p->pos = 0; + p->state = XZ_STATE_STREAM_HEADER; + } + else + { + (*srcLen)++; + src++; + p->padSize++; + } + break; + } + + case XZ_STATE_BLOCK: break; /* to disable GCC warning */ + } + } + /* + if (p->state == XZ_STATE_FINISHED) + *status = CODER_STATUS_FINISHED_WITH_MARK; + return SZ_OK; + */ +} + + +SRes XzUnpacker_CodeFull(CXzUnpacker *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, + ECoderFinishMode finishMode, ECoderStatus *status) +{ + XzUnpacker_Init(p); + XzUnpacker_SetOutBuf(p, dest, *destLen); + + return XzUnpacker_Code(p, + NULL, destLen, + src, srcLen, True, + finishMode, status); +} + + +BoolInt XzUnpacker_IsBlockFinished(const CXzUnpacker *p) +{ + return (p->state == XZ_STATE_BLOCK_HEADER) && (p->pos == 0); +} + +BoolInt XzUnpacker_IsStreamWasFinished(const CXzUnpacker *p) +{ + return (p->state == XZ_STATE_STREAM_PADDING) && (((UInt32)p->padSize & 3) == 0); +} + +UInt64 XzUnpacker_GetExtraSize(const CXzUnpacker *p) +{ + UInt64 num = 0; + if (p->state == XZ_STATE_STREAM_PADDING) + num = p->padSize; + else if (p->state == XZ_STATE_STREAM_HEADER) + num = p->padSize + p->pos; + return num; +} + + + + + + + + + + + + + + + + + + + + + +#ifndef _7ZIP_ST +#include "MtDec.h" +#endif + + +void XzDecMtProps_Init(CXzDecMtProps *p) +{ + p->inBufSize_ST = 1 << 18; + p->outStep_ST = 1 << 20; + p->ignoreErrors = False; + + #ifndef _7ZIP_ST + p->numThreads = 1; + p->inBufSize_MT = 1 << 18; + p->memUseMax = sizeof(size_t) << 28; + #endif +} + + + +#ifndef _7ZIP_ST + +/* ---------- CXzDecMtThread ---------- */ + +typedef struct +{ + Byte *outBuf; + size_t outBufSize; + size_t outPreSize; + size_t inPreSize; + size_t inPreHeaderSize; + size_t blockPackSize_for_Index; // including block header and checksum. + size_t blockPackTotal; // including stream header, block header and checksum. + size_t inCodeSize; + size_t outCodeSize; + ECoderStatus status; + SRes codeRes; + BoolInt skipMode; + // BoolInt finishedWithMark; + EMtDecParseState parseState; + BoolInt parsing_Truncated; + BoolInt atBlockHeader; + CXzStreamFlags streamFlags; + // UInt64 numFinishedStreams + UInt64 numStreams; + UInt64 numTotalBlocks; + UInt64 numBlocks; + + BoolInt dec_created; + CXzUnpacker dec; + + Byte mtPad[1 << 7]; +} CXzDecMtThread; + +#endif + + +/* ---------- CXzDecMt ---------- */ + +typedef struct +{ + CAlignOffsetAlloc alignOffsetAlloc; + ISzAllocPtr allocMid; + + CXzDecMtProps props; + size_t unpackBlockMaxSize; + + ISeqInStream *inStream; + ISeqOutStream *outStream; + ICompressProgress *progress; + + BoolInt finishMode; + BoolInt outSize_Defined; + UInt64 outSize; + + UInt64 outProcessed; + UInt64 inProcessed; + UInt64 readProcessed; + BoolInt readWasFinished; + SRes readRes; + SRes writeRes; + + Byte *outBuf; + size_t outBufSize; + Byte *inBuf; + size_t inBufSize; + + CXzUnpacker dec; + + ECoderStatus status; + SRes codeRes; + + #ifndef _7ZIP_ST + BoolInt mainDecoderWasCalled; + // int statErrorDefined; + int finishedDecoderIndex; + + // global values that are used in Parse stage + CXzStreamFlags streamFlags; + // UInt64 numFinishedStreams + UInt64 numStreams; + UInt64 numTotalBlocks; + UInt64 numBlocks; + + // UInt64 numBadBlocks; + SRes mainErrorCode; // it's set to error code, if the size Code() output doesn't patch the size from Parsing stage + // it can be = SZ_ERROR_INPUT_EOF + // it can be = SZ_ERROR_DATA, in some another cases + BoolInt isBlockHeaderState_Parse; + BoolInt isBlockHeaderState_Write; + UInt64 outProcessed_Parse; + BoolInt parsing_Truncated; + + BoolInt mtc_WasConstructed; + CMtDec mtc; + CXzDecMtThread coders[MTDEC__THREADS_MAX]; + #endif + +} CXzDecMt; + + + +CXzDecMtHandle XzDecMt_Create(ISzAllocPtr alloc, ISzAllocPtr allocMid) +{ + CXzDecMt *p = (CXzDecMt *)ISzAlloc_Alloc(alloc, sizeof(CXzDecMt)); + if (!p) + return NULL; + + AlignOffsetAlloc_CreateVTable(&p->alignOffsetAlloc); + p->alignOffsetAlloc.baseAlloc = alloc; + p->alignOffsetAlloc.numAlignBits = 7; + p->alignOffsetAlloc.offset = 0; + + p->allocMid = allocMid; + + p->outBuf = NULL; + p->outBufSize = 0; + p->inBuf = NULL; + p->inBufSize = 0; + + XzUnpacker_Construct(&p->dec, &p->alignOffsetAlloc.vt); + + p->unpackBlockMaxSize = 0; + + XzDecMtProps_Init(&p->props); + + #ifndef _7ZIP_ST + p->mtc_WasConstructed = False; + { + unsigned i; + for (i = 0; i < MTDEC__THREADS_MAX; i++) + { + CXzDecMtThread *coder = &p->coders[i]; + coder->dec_created = False; + coder->outBuf = NULL; + coder->outBufSize = 0; + } + } + #endif + + return p; +} + + +#ifndef _7ZIP_ST + +static void XzDecMt_FreeOutBufs(CXzDecMt *p) +{ + unsigned i; + for (i = 0; i < MTDEC__THREADS_MAX; i++) + { + CXzDecMtThread *coder = &p->coders[i]; + if (coder->outBuf) + { + ISzAlloc_Free(p->allocMid, coder->outBuf); + coder->outBuf = NULL; + coder->outBufSize = 0; + } + } + p->unpackBlockMaxSize = 0; +} + +#endif + + + +static void XzDecMt_FreeSt(CXzDecMt *p) +{ + XzUnpacker_Free(&p->dec); + + if (p->outBuf) + { + ISzAlloc_Free(p->allocMid, p->outBuf); + p->outBuf = NULL; + } + p->outBufSize = 0; + + if (p->inBuf) + { + ISzAlloc_Free(p->allocMid, p->inBuf); + p->inBuf = NULL; + } + p->inBufSize = 0; +} + + +void XzDecMt_Destroy(CXzDecMtHandle pp) +{ + CXzDecMt *p = (CXzDecMt *)pp; + + XzDecMt_FreeSt(p); + + #ifndef _7ZIP_ST + + if (p->mtc_WasConstructed) + { + MtDec_Destruct(&p->mtc); + p->mtc_WasConstructed = False; + } + { + unsigned i; + for (i = 0; i < MTDEC__THREADS_MAX; i++) + { + CXzDecMtThread *t = &p->coders[i]; + if (t->dec_created) + { + // we don't need to free dict here + XzUnpacker_Free(&t->dec); + t->dec_created = False; + } + } + } + XzDecMt_FreeOutBufs(p); + + #endif + + ISzAlloc_Free(p->alignOffsetAlloc.baseAlloc, pp); +} + + + +#ifndef _7ZIP_ST + +static void XzDecMt_Callback_Parse(void *obj, unsigned coderIndex, CMtDecCallbackInfo *cc) +{ + CXzDecMt *me = (CXzDecMt *)obj; + CXzDecMtThread *coder = &me->coders[coderIndex]; + size_t srcSize = cc->srcSize; + + cc->srcSize = 0; + cc->outPos = 0; + cc->state = MTDEC_PARSE_CONTINUE; + + cc->canCreateNewThread = True; + + if (cc->startCall) + { + coder->outPreSize = 0; + coder->inPreSize = 0; + coder->inPreHeaderSize = 0; + coder->parseState = MTDEC_PARSE_CONTINUE; + coder->parsing_Truncated = False; + coder->skipMode = False; + coder->codeRes = SZ_OK; + coder->status = CODER_STATUS_NOT_SPECIFIED; + coder->inCodeSize = 0; + coder->outCodeSize = 0; + + coder->numStreams = me->numStreams; + coder->numTotalBlocks = me->numTotalBlocks; + coder->numBlocks = me->numBlocks; + + if (!coder->dec_created) + { + XzUnpacker_Construct(&coder->dec, &me->alignOffsetAlloc.vt); + coder->dec_created = True; + } + + XzUnpacker_Init(&coder->dec); + + if (me->isBlockHeaderState_Parse) + { + coder->dec.streamFlags = me->streamFlags; + coder->atBlockHeader = True; + XzUnpacker_PrepareToRandomBlockDecoding(&coder->dec); + } + else + { + coder->atBlockHeader = False; + me->isBlockHeaderState_Parse = True; + } + + coder->dec.numStartedStreams = me->numStreams; + coder->dec.numTotalBlocks = me->numTotalBlocks; + coder->dec.numBlocks = me->numBlocks; + } + + while (!coder->skipMode) + { + ECoderStatus status; + SRes res; + size_t srcSize2 = srcSize; + size_t destSize = (size_t)0 - 1; + + coder->dec.parseMode = True; + coder->dec.headerParsedOk = False; + + PRF_STR_INT("Parse", srcSize2); + + res = XzUnpacker_Code(&coder->dec, + NULL, &destSize, + cc->src, &srcSize2, cc->srcFinished, + CODER_FINISH_END, &status); + + // PRF(printf(" res = %d, srcSize2 = %d", res, (unsigned)srcSize2)); + + coder->codeRes = res; + coder->status = status; + cc->srcSize += srcSize2; + srcSize -= srcSize2; + coder->inPreHeaderSize += srcSize2; + coder->inPreSize = coder->inPreHeaderSize; + + if (res != SZ_OK) + { + cc->state = + coder->parseState = MTDEC_PARSE_END; + /* + if (res == SZ_ERROR_MEM) + return res; + return SZ_OK; + */ + return; // res; + } + + if (coder->dec.headerParsedOk) + { + const CXzBlock *block = &coder->dec.block; + if (XzBlock_HasUnpackSize(block) + // && block->unpackSize <= me->props.outBlockMax + && XzBlock_HasPackSize(block)) + { + { + if (block->unpackSize * 2 * me->mtc.numStartedThreads > me->props.memUseMax) + { + cc->state = MTDEC_PARSE_OVERFLOW; + return; // SZ_OK; + } + } + { + UInt64 packSize = block->packSize; + UInt64 packSizeAligned = packSize + ((0 - (unsigned)packSize) & 3); + UInt32 checkSize = XzFlags_GetCheckSize(coder->dec.streamFlags); + UInt64 blockPackSum = coder->inPreSize + packSizeAligned + checkSize; + // if (blockPackSum <= me->props.inBlockMax) + // unpackBlockMaxSize + { + coder->blockPackSize_for_Index = (size_t)(coder->dec.blockHeaderSize + packSize + checkSize); + coder->blockPackTotal = (size_t)blockPackSum; + coder->outPreSize = (size_t)block->unpackSize; + coder->streamFlags = coder->dec.streamFlags; + me->streamFlags = coder->dec.streamFlags; + coder->skipMode = True; + break; + } + } + } + } + else + // if (coder->inPreSize <= me->props.inBlockMax) + { + if (!cc->srcFinished) + return; // SZ_OK; + cc->state = + coder->parseState = MTDEC_PARSE_END; + return; // SZ_OK; + } + cc->state = MTDEC_PARSE_OVERFLOW; + return; // SZ_OK; + } + + // ---------- skipMode ---------- + { + UInt64 rem = coder->blockPackTotal - coder->inPreSize; + size_t cur = srcSize; + if (cur > rem) + cur = (size_t)rem; + cc->srcSize += cur; + coder->inPreSize += cur; + srcSize -= cur; + + if (coder->inPreSize == coder->blockPackTotal) + { + if (srcSize == 0) + { + if (!cc->srcFinished) + return; // SZ_OK; + cc->state = MTDEC_PARSE_END; + } + else if ((cc->src)[cc->srcSize] == 0) // we check control byte of next block + cc->state = MTDEC_PARSE_END; + else + { + cc->state = MTDEC_PARSE_NEW; + + { + size_t blockMax = me->unpackBlockMaxSize; + if (blockMax < coder->outPreSize) + blockMax = coder->outPreSize; + { + UInt64 required = (UInt64)blockMax * (me->mtc.numStartedThreads + 1) * 2; + if (me->props.memUseMax < required) + cc->canCreateNewThread = False; + } + } + + if (me->outSize_Defined) + { + // next block can be zero size + const UInt64 rem2 = me->outSize - me->outProcessed_Parse; + if (rem2 < coder->outPreSize) + { + coder->parsing_Truncated = True; + cc->state = MTDEC_PARSE_END; + } + me->outProcessed_Parse += coder->outPreSize; + } + } + } + else if (cc->srcFinished) + cc->state = MTDEC_PARSE_END; + else + return; // SZ_OK; + + coder->parseState = cc->state; + cc->outPos = coder->outPreSize; + + me->numStreams = coder->dec.numStartedStreams; + me->numTotalBlocks = coder->dec.numTotalBlocks; + me->numBlocks = coder->dec.numBlocks + 1; + return; // SZ_OK; + } +} + + +static SRes XzDecMt_Callback_PreCode(void *pp, unsigned coderIndex) +{ + CXzDecMt *me = (CXzDecMt *)pp; + CXzDecMtThread *coder = &me->coders[coderIndex]; + Byte *dest; + + if (!coder->dec.headerParsedOk) + return SZ_OK; + + dest = coder->outBuf; + + if (!dest || coder->outBufSize < coder->outPreSize) + { + if (dest) + { + ISzAlloc_Free(me->allocMid, dest); + coder->outBuf = NULL; + coder->outBufSize = 0; + } + { + size_t outPreSize = coder->outPreSize; + if (outPreSize == 0) + outPreSize = 1; + dest = (Byte *)ISzAlloc_Alloc(me->allocMid, outPreSize); + } + if (!dest) + return SZ_ERROR_MEM; + coder->outBuf = dest; + coder->outBufSize = coder->outPreSize; + + if (coder->outBufSize > me->unpackBlockMaxSize) + me->unpackBlockMaxSize = coder->outBufSize; + } + + // return SZ_ERROR_MEM; + + XzUnpacker_SetOutBuf(&coder->dec, coder->outBuf, coder->outBufSize); + + { + SRes res = XzDecMix_Init(&coder->dec.decoder, &coder->dec.block, coder->outBuf, coder->outBufSize); + // res = SZ_ERROR_UNSUPPORTED; // to test + coder->codeRes = res; + if (res != SZ_OK) + { + // if (res == SZ_ERROR_MEM) return res; + if (me->props.ignoreErrors && res != SZ_ERROR_MEM) + return SZ_OK; + return res; + } + } + + return SZ_OK; +} + + +static SRes XzDecMt_Callback_Code(void *pp, unsigned coderIndex, + const Byte *src, size_t srcSize, int srcFinished, + // int finished, int blockFinished, + UInt64 *inCodePos, UInt64 *outCodePos, int *stop) +{ + CXzDecMt *me = (CXzDecMt *)pp; + CXzDecMtThread *coder = &me->coders[coderIndex]; + + *inCodePos = coder->inCodeSize; + *outCodePos = coder->outCodeSize; + *stop = True; + + if (srcSize > coder->inPreSize - coder->inCodeSize) + return SZ_ERROR_FAIL; + + if (coder->inCodeSize < coder->inPreHeaderSize) + { + size_t step = coder->inPreHeaderSize - coder->inCodeSize; + if (step > srcSize) + step = srcSize; + src += step; + srcSize -= step; + coder->inCodeSize += step; + *inCodePos = coder->inCodeSize; + if (coder->inCodeSize < coder->inPreHeaderSize) + { + *stop = False; + return SZ_OK; + } + } + + if (!coder->dec.headerParsedOk) + return SZ_OK; + if (!coder->outBuf) + return SZ_OK; + + if (coder->codeRes == SZ_OK) + { + ECoderStatus status; + SRes res; + size_t srcProcessed = srcSize; + size_t outSizeCur = coder->outPreSize - coder->dec.outDataWritten; + + // PRF(printf("\nCallback_Code: Code %d %d\n", (unsigned)srcSize, (unsigned)outSizeCur)); + + res = XzUnpacker_Code(&coder->dec, + NULL, &outSizeCur, + src, &srcProcessed, srcFinished, + // coder->finishedWithMark ? CODER_FINISH_END : CODER_FINISH_ANY, + CODER_FINISH_END, + &status); + + // PRF(printf(" res = %d, srcSize2 = %d, outSizeCur = %d", res, (unsigned)srcProcessed, (unsigned)outSizeCur)); + + coder->codeRes = res; + coder->status = status; + coder->inCodeSize += srcProcessed; + coder->outCodeSize = coder->dec.outDataWritten; + *inCodePos = coder->inCodeSize; + *outCodePos = coder->outCodeSize; + + if (res == SZ_OK) + { + if (srcProcessed == srcSize) + *stop = False; + return SZ_OK; + } + } + + if (me->props.ignoreErrors && coder->codeRes != SZ_ERROR_MEM) + { + *inCodePos = coder->inPreSize; + *outCodePos = coder->outPreSize; + return SZ_OK; + } + return coder->codeRes; +} + + +#define XZDECMT_STREAM_WRITE_STEP (1 << 24) + +static SRes XzDecMt_Callback_Write(void *pp, unsigned coderIndex, + BoolInt needWriteToStream, + const Byte *src, size_t srcSize, BoolInt isCross, + // int srcFinished, + BoolInt *needContinue, + BoolInt *canRecode) +{ + CXzDecMt *me = (CXzDecMt *)pp; + const CXzDecMtThread *coder = &me->coders[coderIndex]; + + // PRF(printf("\nWrite processed = %d srcSize = %d\n", (unsigned)me->mtc.inProcessed, (unsigned)srcSize)); + + *needContinue = False; + *canRecode = True; + + if (!needWriteToStream) + return SZ_OK; + + if (!coder->dec.headerParsedOk || !coder->outBuf) + { + if (me->finishedDecoderIndex < 0) + me->finishedDecoderIndex = (int)coderIndex; + return SZ_OK; + } + + if (me->finishedDecoderIndex >= 0) + return SZ_OK; + + me->mtc.inProcessed += coder->inCodeSize; + + *canRecode = False; + + { + SRes res; + size_t size = coder->outCodeSize; + Byte *data = coder->outBuf; + + // we use in me->dec: sha, numBlocks, indexSize + + if (!me->isBlockHeaderState_Write) + { + XzUnpacker_PrepareToRandomBlockDecoding(&me->dec); + me->dec.decodeOnlyOneBlock = False; + me->dec.numStartedStreams = coder->dec.numStartedStreams; + me->dec.streamFlags = coder->streamFlags; + + me->isBlockHeaderState_Write = True; + } + + me->dec.numTotalBlocks = coder->dec.numTotalBlocks; + XzUnpacker_UpdateIndex(&me->dec, coder->blockPackSize_for_Index, coder->outPreSize); + + if (coder->outPreSize != size) + { + if (me->props.ignoreErrors) + { + memset(data + size, 0, coder->outPreSize - size); + size = coder->outPreSize; + } + // me->numBadBlocks++; + if (me->mainErrorCode == SZ_OK) + { + if ((int)coder->status == LZMA_STATUS_NEEDS_MORE_INPUT) + me->mainErrorCode = SZ_ERROR_INPUT_EOF; + else + me->mainErrorCode = SZ_ERROR_DATA; + } + } + + if (me->writeRes != SZ_OK) + return me->writeRes; + + res = SZ_OK; + { + if (me->outSize_Defined) + { + const UInt64 rem = me->outSize - me->outProcessed; + if (size > rem) + size = (SizeT)rem; + } + + for (;;) + { + size_t cur = size; + size_t written; + if (cur > XZDECMT_STREAM_WRITE_STEP) + cur = XZDECMT_STREAM_WRITE_STEP; + + written = ISeqOutStream_Write(me->outStream, data, cur); + + // PRF(printf("\nWritten ask = %d written = %d\n", (unsigned)cur, (unsigned)written)); + + me->outProcessed += written; + if (written != cur) + { + me->writeRes = SZ_ERROR_WRITE; + res = me->writeRes; + break; + } + data += cur; + size -= cur; + // PRF_STR_INT("Written size =", size); + if (size == 0) + break; + res = MtProgress_ProgressAdd(&me->mtc.mtProgress, 0, 0); + if (res != SZ_OK) + break; + } + } + + if (coder->codeRes != SZ_OK) + if (!me->props.ignoreErrors) + { + me->finishedDecoderIndex = (int)coderIndex; + return res; + } + + RINOK(res); + + if (coder->inPreSize != coder->inCodeSize + || coder->blockPackTotal != coder->inCodeSize) + { + me->finishedDecoderIndex = (int)coderIndex; + return SZ_OK; + } + + if (coder->parseState != MTDEC_PARSE_END) + { + *needContinue = True; + return SZ_OK; + } + } + + // (coder->state == MTDEC_PARSE_END) means that there are no other working threads + // so we can use mtc variables without lock + + PRF_STR_INT("Write MTDEC_PARSE_END", me->mtc.inProcessed); + + me->mtc.mtProgress.totalInSize = me->mtc.inProcessed; + { + CXzUnpacker *dec = &me->dec; + + PRF_STR_INT("PostSingle", srcSize); + + { + size_t srcProcessed = srcSize; + ECoderStatus status; + size_t outSizeCur = 0; + SRes res; + + // dec->decodeOnlyOneBlock = False; + dec->decodeToStreamSignature = True; + + me->mainDecoderWasCalled = True; + + if (coder->parsing_Truncated) + { + me->parsing_Truncated = True; + return SZ_OK; + } + + /* + We have processed all xz-blocks of stream, + And xz unpacker is at XZ_STATE_BLOCK_HEADER state, where + (src) is a pointer to xz-Index structure. + We finish reading of current xz-Stream, including Zero padding after xz-Stream. + We exit, if we reach extra byte (first byte of new-Stream or another data). + But we don't update input stream pointer for that new extra byte. + If extra byte is not correct first byte of xz-signature, + we have SZ_ERROR_NO_ARCHIVE error here. + */ + + res = XzUnpacker_Code(dec, + NULL, &outSizeCur, + src, &srcProcessed, + me->mtc.readWasFinished, // srcFinished + CODER_FINISH_END, // CODER_FINISH_ANY, + &status); + + // res = SZ_ERROR_ARCHIVE; // for failure test + + me->status = status; + me->codeRes = res; + + if (isCross) + me->mtc.crossStart += srcProcessed; + + me->mtc.inProcessed += srcProcessed; + me->mtc.mtProgress.totalInSize = me->mtc.inProcessed; + + srcSize -= srcProcessed; + src += srcProcessed; + + if (res != SZ_OK) + { + return SZ_OK; + // return res; + } + + if (dec->state == XZ_STATE_STREAM_HEADER) + { + *needContinue = True; + me->isBlockHeaderState_Parse = False; + me->isBlockHeaderState_Write = False; + + if (!isCross) + { + Byte *crossBuf = MtDec_GetCrossBuff(&me->mtc); + if (!crossBuf) + return SZ_ERROR_MEM; + if (srcSize != 0) + memcpy(crossBuf, src, srcSize); + me->mtc.crossStart = 0; + me->mtc.crossEnd = srcSize; + } + + PRF_STR_INT("XZ_STATE_STREAM_HEADER crossEnd = ", (unsigned)me->mtc.crossEnd); + + return SZ_OK; + } + + if (status != CODER_STATUS_NEEDS_MORE_INPUT || srcSize != 0) + { + return SZ_ERROR_FAIL; + } + + if (me->mtc.readWasFinished) + { + return SZ_OK; + } + } + + { + size_t inPos; + size_t inLim; + // const Byte *inData; + UInt64 inProgressPrev = me->mtc.inProcessed; + + // XzDecMt_Prepare_InBuf_ST(p); + Byte *crossBuf = MtDec_GetCrossBuff(&me->mtc); + if (!crossBuf) + return SZ_ERROR_MEM; + + inPos = 0; + inLim = 0; + + // inData = crossBuf; + + for (;;) + { + SizeT inProcessed; + SizeT outProcessed; + ECoderStatus status; + SRes res; + + if (inPos == inLim) + { + if (!me->mtc.readWasFinished) + { + inPos = 0; + inLim = me->mtc.inBufSize; + me->mtc.readRes = ISeqInStream_Read(me->inStream, (void *)crossBuf, &inLim); + me->mtc.readProcessed += inLim; + if (inLim == 0 || me->mtc.readRes != SZ_OK) + me->mtc.readWasFinished = True; + } + } + + inProcessed = inLim - inPos; + outProcessed = 0; + + res = XzUnpacker_Code(dec, + NULL, &outProcessed, + crossBuf + inPos, &inProcessed, + (inProcessed == 0), // srcFinished + CODER_FINISH_END, &status); + + me->codeRes = res; + me->status = status; + inPos += inProcessed; + me->mtc.inProcessed += inProcessed; + me->mtc.mtProgress.totalInSize = me->mtc.inProcessed; + + if (res != SZ_OK) + { + return SZ_OK; + // return res; + } + + if (dec->state == XZ_STATE_STREAM_HEADER) + { + *needContinue = True; + me->mtc.crossStart = inPos; + me->mtc.crossEnd = inLim; + me->isBlockHeaderState_Parse = False; + me->isBlockHeaderState_Write = False; + return SZ_OK; + } + + if (status != CODER_STATUS_NEEDS_MORE_INPUT) + return SZ_ERROR_FAIL; + + if (me->mtc.progress) + { + UInt64 inDelta = me->mtc.inProcessed - inProgressPrev; + if (inDelta >= (1 << 22)) + { + RINOK(MtProgress_Progress_ST(&me->mtc.mtProgress)); + inProgressPrev = me->mtc.inProcessed; + } + } + if (me->mtc.readWasFinished) + return SZ_OK; + } + } + } +} + + +#endif + + + +void XzStatInfo_Clear(CXzStatInfo *p) +{ + p->InSize = 0; + p->OutSize = 0; + + p->NumStreams = 0; + p->NumBlocks = 0; + + p->UnpackSize_Defined = False; + + p->NumStreams_Defined = False; + p->NumBlocks_Defined = False; + + p->DataAfterEnd = False; + p->DecodingTruncated = False; + + p->DecodeRes = SZ_OK; + p->ReadRes = SZ_OK; + p->ProgressRes = SZ_OK; + + p->CombinedRes = SZ_OK; + p->CombinedRes_Type = SZ_OK; +} + + + +/* + XzDecMt_Decode_ST() can return SZ_OK or the following errors + - SZ_ERROR_MEM for memory allocation error + - error from XzUnpacker_Code() function + - SZ_ERROR_WRITE for ISeqOutStream::Write(). stat->CombinedRes_Type = SZ_ERROR_WRITE in that case + - ICompressProgress::Progress() error, stat->CombinedRes_Type = SZ_ERROR_PROGRESS. + But XzDecMt_Decode_ST() doesn't return ISeqInStream::Read() errors. + ISeqInStream::Read() result is set to p->readRes. + also it can set stat->CombinedRes_Type to SZ_ERROR_WRITE or SZ_ERROR_PROGRESS. +*/ + +static SRes XzDecMt_Decode_ST(CXzDecMt *p + #ifndef _7ZIP_ST + , BoolInt tMode + #endif + , CXzStatInfo *stat) +{ + size_t outPos; + size_t inPos, inLim; + const Byte *inData; + UInt64 inPrev, outPrev; + + CXzUnpacker *dec; + + #ifndef _7ZIP_ST + if (tMode) + { + XzDecMt_FreeOutBufs(p); + tMode = MtDec_PrepareRead(&p->mtc); + } + #endif + + if (!p->outBuf || p->outBufSize != p->props.outStep_ST) + { + ISzAlloc_Free(p->allocMid, p->outBuf); + p->outBufSize = 0; + p->outBuf = (Byte *)ISzAlloc_Alloc(p->allocMid, p->props.outStep_ST); + if (!p->outBuf) + return SZ_ERROR_MEM; + p->outBufSize = p->props.outStep_ST; + } + + if (!p->inBuf || p->inBufSize != p->props.inBufSize_ST) + { + ISzAlloc_Free(p->allocMid, p->inBuf); + p->inBufSize = 0; + p->inBuf = (Byte *)ISzAlloc_Alloc(p->allocMid, p->props.inBufSize_ST); + if (!p->inBuf) + return SZ_ERROR_MEM; + p->inBufSize = p->props.inBufSize_ST; + } + + dec = &p->dec; + dec->decodeToStreamSignature = False; + // dec->decodeOnlyOneBlock = False; + + XzUnpacker_SetOutBuf(dec, NULL, 0); + + inPrev = p->inProcessed; + outPrev = p->outProcessed; + + inPos = 0; + inLim = 0; + inData = NULL; + outPos = 0; + + for (;;) + { + SizeT outSize; + BoolInt finished; + ECoderFinishMode finishMode; + SizeT inProcessed; + ECoderStatus status; + SRes res; + + SizeT outProcessed; + + + + if (inPos == inLim) + { + #ifndef _7ZIP_ST + if (tMode) + { + inData = MtDec_Read(&p->mtc, &inLim); + inPos = 0; + if (inData) + continue; + tMode = False; + inLim = 0; + } + #endif + + if (!p->readWasFinished) + { + inPos = 0; + inLim = p->inBufSize; + inData = p->inBuf; + p->readRes = ISeqInStream_Read(p->inStream, (void *)p->inBuf, &inLim); + p->readProcessed += inLim; + if (inLim == 0 || p->readRes != SZ_OK) + p->readWasFinished = True; + } + } + + outSize = p->props.outStep_ST - outPos; + + finishMode = CODER_FINISH_ANY; + if (p->outSize_Defined) + { + const UInt64 rem = p->outSize - p->outProcessed; + if (outSize >= rem) + { + outSize = (SizeT)rem; + if (p->finishMode) + finishMode = CODER_FINISH_END; + } + } + + inProcessed = inLim - inPos; + outProcessed = outSize; + + res = XzUnpacker_Code(dec, p->outBuf + outPos, &outProcessed, + inData + inPos, &inProcessed, + (inPos == inLim), // srcFinished + finishMode, &status); + + p->codeRes = res; + p->status = status; + + inPos += inProcessed; + outPos += outProcessed; + p->inProcessed += inProcessed; + p->outProcessed += outProcessed; + + finished = ((inProcessed == 0 && outProcessed == 0) || res != SZ_OK); + + if (finished || outProcessed >= outSize) + if (outPos != 0) + { + const size_t written = ISeqOutStream_Write(p->outStream, p->outBuf, outPos); + // p->outProcessed += written; // 21.01: BUG fixed + if (written != outPos) + { + stat->CombinedRes_Type = SZ_ERROR_WRITE; + return SZ_ERROR_WRITE; + } + outPos = 0; + } + + if (p->progress && res == SZ_OK) + { + if (p->inProcessed - inPrev >= (1 << 22) || + p->outProcessed - outPrev >= (1 << 22)) + { + res = ICompressProgress_Progress(p->progress, p->inProcessed, p->outProcessed); + if (res != SZ_OK) + { + stat->CombinedRes_Type = SZ_ERROR_PROGRESS; + stat->ProgressRes = res; + return res; + } + inPrev = p->inProcessed; + outPrev = p->outProcessed; + } + } + + if (finished) + { + // p->codeRes is preliminary error from XzUnpacker_Code. + // and it can be corrected later as final result + // so we return SZ_OK here instead of (res); + return SZ_OK; + // return res; + } + } +} + + + +/* +XzStatInfo_SetStat() transforms + CXzUnpacker return code and status to combined CXzStatInfo results. + it can convert SZ_OK to SZ_ERROR_INPUT_EOF + it can convert SZ_ERROR_NO_ARCHIVE to SZ_OK and (DataAfterEnd = 1) +*/ + +static void XzStatInfo_SetStat(const CXzUnpacker *dec, + int finishMode, + // UInt64 readProcessed, + UInt64 inProcessed, + SRes res, // it's result from CXzUnpacker unpacker + ECoderStatus status, + BoolInt decodingTruncated, + CXzStatInfo *stat) +{ + UInt64 extraSize; + + stat->DecodingTruncated = (Byte)(decodingTruncated ? 1 : 0); + stat->InSize = inProcessed; + stat->NumStreams = dec->numStartedStreams; + stat->NumBlocks = dec->numTotalBlocks; + + stat->UnpackSize_Defined = True; + stat->NumStreams_Defined = True; + stat->NumBlocks_Defined = True; + + extraSize = XzUnpacker_GetExtraSize(dec); + + if (res == SZ_OK) + { + if (status == CODER_STATUS_NEEDS_MORE_INPUT) + { + // CODER_STATUS_NEEDS_MORE_INPUT is expected status for correct xz streams + // any extra data is part of correct data + extraSize = 0; + // if xz stream was not finished, then we need more data + if (!XzUnpacker_IsStreamWasFinished(dec)) + res = SZ_ERROR_INPUT_EOF; + } + else + { + // CODER_STATUS_FINISHED_WITH_MARK is not possible for multi stream xz decoding + // so he we have (status == CODER_STATUS_NOT_FINISHED) + // if (status != CODER_STATUS_FINISHED_WITH_MARK) + if (!decodingTruncated || finishMode) + res = SZ_ERROR_DATA; + } + } + else if (res == SZ_ERROR_NO_ARCHIVE) + { + /* + SZ_ERROR_NO_ARCHIVE is possible for 2 states: + XZ_STATE_STREAM_HEADER - if bad signature or bad CRC + XZ_STATE_STREAM_PADDING - if non-zero padding data + extraSize and inProcessed don't include "bad" byte + */ + // if (inProcessed == extraSize), there was no any good xz stream header, and we keep error + if (inProcessed != extraSize) // if there were good xz streams before error + { + // if (extraSize != 0 || readProcessed != inProcessed) + { + // he we suppose that all xz streams were finsihed OK, and we have + // some extra data after all streams + stat->DataAfterEnd = True; + res = SZ_OK; + } + } + } + + if (stat->DecodeRes == SZ_OK) + stat->DecodeRes = res; + + stat->InSize -= extraSize; +} + + + +SRes XzDecMt_Decode(CXzDecMtHandle pp, + const CXzDecMtProps *props, + const UInt64 *outDataSize, int finishMode, + ISeqOutStream *outStream, + // Byte *outBuf, size_t *outBufSize, + ISeqInStream *inStream, + // const Byte *inData, size_t inDataSize, + CXzStatInfo *stat, + int *isMT, + ICompressProgress *progress) +{ + CXzDecMt *p = (CXzDecMt *)pp; + #ifndef _7ZIP_ST + BoolInt tMode; + #endif + + XzStatInfo_Clear(stat); + + p->props = *props; + + p->inStream = inStream; + p->outStream = outStream; + p->progress = progress; + // p->stat = stat; + + p->outSize = 0; + p->outSize_Defined = False; + if (outDataSize) + { + p->outSize_Defined = True; + p->outSize = *outDataSize; + } + + p->finishMode = finishMode; + + // p->outSize = 457; p->outSize_Defined = True; p->finishMode = False; // for test + + p->writeRes = SZ_OK; + p->outProcessed = 0; + p->inProcessed = 0; + p->readProcessed = 0; + p->readWasFinished = False; + p->readRes = SZ_OK; + + p->codeRes = SZ_OK; + p->status = CODER_STATUS_NOT_SPECIFIED; + + XzUnpacker_Init(&p->dec); + + *isMT = False; + + /* + p->outBuf = NULL; + p->outBufSize = 0; + if (!outStream) + { + p->outBuf = outBuf; + p->outBufSize = *outBufSize; + *outBufSize = 0; + } + */ + + + #ifndef _7ZIP_ST + + p->isBlockHeaderState_Parse = False; + p->isBlockHeaderState_Write = False; + // p->numBadBlocks = 0; + p->mainErrorCode = SZ_OK; + p->mainDecoderWasCalled = False; + + tMode = False; + + if (p->props.numThreads > 1) + { + IMtDecCallback2 vt; + BoolInt needContinue; + SRes res; + // we just free ST buffers here + // but we still keep state variables, that was set in XzUnpacker_Init() + XzDecMt_FreeSt(p); + + p->outProcessed_Parse = 0; + p->parsing_Truncated = False; + + p->numStreams = 0; + p->numTotalBlocks = 0; + p->numBlocks = 0; + p->finishedDecoderIndex = -1; + + if (!p->mtc_WasConstructed) + { + p->mtc_WasConstructed = True; + MtDec_Construct(&p->mtc); + } + + p->mtc.mtCallback = &vt; + p->mtc.mtCallbackObject = p; + + p->mtc.progress = progress; + p->mtc.inStream = inStream; + p->mtc.alloc = &p->alignOffsetAlloc.vt; + // p->mtc.inData = inData; + // p->mtc.inDataSize = inDataSize; + p->mtc.inBufSize = p->props.inBufSize_MT; + // p->mtc.inBlockMax = p->props.inBlockMax; + p->mtc.numThreadsMax = p->props.numThreads; + + *isMT = True; + + vt.Parse = XzDecMt_Callback_Parse; + vt.PreCode = XzDecMt_Callback_PreCode; + vt.Code = XzDecMt_Callback_Code; + vt.Write = XzDecMt_Callback_Write; + + + res = MtDec_Code(&p->mtc); + + + stat->InSize = p->mtc.inProcessed; + + p->inProcessed = p->mtc.inProcessed; + p->readRes = p->mtc.readRes; + p->readWasFinished = p->mtc.readWasFinished; + p->readProcessed = p->mtc.readProcessed; + + tMode = True; + needContinue = False; + + if (res == SZ_OK) + { + if (p->mtc.mtProgress.res != SZ_OK) + { + res = p->mtc.mtProgress.res; + stat->ProgressRes = res; + stat->CombinedRes_Type = SZ_ERROR_PROGRESS; + } + else + needContinue = p->mtc.needContinue; + } + + if (!needContinue) + { + { + SRes codeRes; + BoolInt truncated = False; + ECoderStatus status; + const CXzUnpacker *dec; + + stat->OutSize = p->outProcessed; + + if (p->finishedDecoderIndex >= 0) + { + const CXzDecMtThread *coder = &p->coders[(unsigned)p->finishedDecoderIndex]; + codeRes = coder->codeRes; + dec = &coder->dec; + status = coder->status; + } + else if (p->mainDecoderWasCalled) + { + codeRes = p->codeRes; + dec = &p->dec; + status = p->status; + truncated = p->parsing_Truncated; + } + else + return SZ_ERROR_FAIL; + + if (p->mainErrorCode != SZ_OK) + stat->DecodeRes = p->mainErrorCode; + + XzStatInfo_SetStat(dec, p->finishMode, + // p->mtc.readProcessed, + p->mtc.inProcessed, + codeRes, status, + truncated, + stat); + } + + if (res == SZ_OK) + { + stat->ReadRes = p->mtc.readRes; + + if (p->writeRes != SZ_OK) + { + res = p->writeRes; + stat->CombinedRes_Type = SZ_ERROR_WRITE; + } + else if (p->mtc.readRes != SZ_OK + // && p->mtc.inProcessed == p->mtc.readProcessed + && stat->DecodeRes == SZ_ERROR_INPUT_EOF) + { + res = p->mtc.readRes; + stat->CombinedRes_Type = SZ_ERROR_READ; + } + else if (stat->DecodeRes != SZ_OK) + res = stat->DecodeRes; + } + + stat->CombinedRes = res; + if (stat->CombinedRes_Type == SZ_OK) + stat->CombinedRes_Type = res; + return res; + } + + PRF_STR("----- decoding ST -----"); + } + + #endif + + + *isMT = False; + + { + SRes res = XzDecMt_Decode_ST(p + #ifndef _7ZIP_ST + , tMode + #endif + , stat + ); + + #ifndef _7ZIP_ST + // we must set error code from MT decoding at first + if (p->mainErrorCode != SZ_OK) + stat->DecodeRes = p->mainErrorCode; + #endif + + XzStatInfo_SetStat(&p->dec, + p->finishMode, + // p->readProcessed, + p->inProcessed, + p->codeRes, p->status, + False, // truncated + stat); + + stat->ReadRes = p->readRes; + + if (res == SZ_OK) + { + if (p->readRes != SZ_OK + // && p->inProcessed == p->readProcessed + && stat->DecodeRes == SZ_ERROR_INPUT_EOF) + { + // we set read error as combined error, only if that error was the reason + // of decoding problem + res = p->readRes; + stat->CombinedRes_Type = SZ_ERROR_READ; + } + else if (stat->DecodeRes != SZ_OK) + res = stat->DecodeRes; + } + + stat->CombinedRes = res; + if (stat->CombinedRes_Type == SZ_OK) + stat->CombinedRes_Type = res; + return res; + } +} diff --git a/deps/LZMA-SDK/C/XzEnc.c b/deps/LZMA-SDK/C/XzEnc.c index 759ba670e..be174cccf 100644 --- a/deps/LZMA-SDK/C/XzEnc.c +++ b/deps/LZMA-SDK/C/XzEnc.c @@ -1,1330 +1,1330 @@ -/* XzEnc.c -- Xz Encode -2021-04-01 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include -#include - -#include "7zCrc.h" -#include "Bra.h" -#include "CpuArch.h" - -#ifdef USE_SUBBLOCK -#include "Bcj3Enc.c" -#include "SbFind.c" -#include "SbEnc.c" -#endif - -#include "XzEnc.h" - -// #define _7ZIP_ST - -#ifndef _7ZIP_ST -#include "MtCoder.h" -#else -#define MTCODER__THREADS_MAX 1 -#define MTCODER__BLOCKS_MAX 1 -#endif - -#define XZ_GET_PAD_SIZE(dataSize) ((4 - ((unsigned)(dataSize) & 3)) & 3) - -/* max pack size for LZMA2 block + check-64bytrs: */ -#define XZ_GET_MAX_BLOCK_PACK_SIZE(unpackSize) ((unpackSize) + ((unpackSize) >> 10) + 16 + 64) - -#define XZ_GET_ESTIMATED_BLOCK_TOTAL_PACK_SIZE(unpackSize) (XZ_BLOCK_HEADER_SIZE_MAX + XZ_GET_MAX_BLOCK_PACK_SIZE(unpackSize)) - - -#define XzBlock_ClearFlags(p) (p)->flags = 0; -#define XzBlock_SetNumFilters(p, n) (p)->flags = (Byte)((p)->flags | ((n) - 1)); -#define XzBlock_SetHasPackSize(p) (p)->flags |= XZ_BF_PACK_SIZE; -#define XzBlock_SetHasUnpackSize(p) (p)->flags |= XZ_BF_UNPACK_SIZE; - - -static SRes WriteBytes(ISeqOutStream *s, const void *buf, size_t size) -{ - return (ISeqOutStream_Write(s, buf, size) == size) ? SZ_OK : SZ_ERROR_WRITE; -} - -static SRes WriteBytesUpdateCrc(ISeqOutStream *s, const void *buf, size_t size, UInt32 *crc) -{ - *crc = CrcUpdate(*crc, buf, size); - return WriteBytes(s, buf, size); -} - - -static SRes Xz_WriteHeader(CXzStreamFlags f, ISeqOutStream *s) -{ - UInt32 crc; - Byte header[XZ_STREAM_HEADER_SIZE]; - memcpy(header, XZ_SIG, XZ_SIG_SIZE); - header[XZ_SIG_SIZE] = (Byte)(f >> 8); - header[XZ_SIG_SIZE + 1] = (Byte)(f & 0xFF); - crc = CrcCalc(header + XZ_SIG_SIZE, XZ_STREAM_FLAGS_SIZE); - SetUi32(header + XZ_SIG_SIZE + XZ_STREAM_FLAGS_SIZE, crc); - return WriteBytes(s, header, XZ_STREAM_HEADER_SIZE); -} - - -static SRes XzBlock_WriteHeader(const CXzBlock *p, ISeqOutStream *s) -{ - Byte header[XZ_BLOCK_HEADER_SIZE_MAX]; - - unsigned pos = 1; - unsigned numFilters, i; - header[pos++] = p->flags; - - if (XzBlock_HasPackSize(p)) pos += Xz_WriteVarInt(header + pos, p->packSize); - if (XzBlock_HasUnpackSize(p)) pos += Xz_WriteVarInt(header + pos, p->unpackSize); - numFilters = XzBlock_GetNumFilters(p); - - for (i = 0; i < numFilters; i++) - { - const CXzFilter *f = &p->filters[i]; - pos += Xz_WriteVarInt(header + pos, f->id); - pos += Xz_WriteVarInt(header + pos, f->propsSize); - memcpy(header + pos, f->props, f->propsSize); - pos += f->propsSize; - } - - while ((pos & 3) != 0) - header[pos++] = 0; - - header[0] = (Byte)(pos >> 2); - SetUi32(header + pos, CrcCalc(header, pos)); - return WriteBytes(s, header, pos + 4); -} - - - - -typedef struct -{ - size_t numBlocks; - size_t size; - size_t allocated; - Byte *blocks; -} CXzEncIndex; - - -static void XzEncIndex_Construct(CXzEncIndex *p) -{ - p->numBlocks = 0; - p->size = 0; - p->allocated = 0; - p->blocks = NULL; -} - -static void XzEncIndex_Init(CXzEncIndex *p) -{ - p->numBlocks = 0; - p->size = 0; -} - -static void XzEncIndex_Free(CXzEncIndex *p, ISzAllocPtr alloc) -{ - if (p->blocks) - { - ISzAlloc_Free(alloc, p->blocks); - p->blocks = NULL; - } - p->numBlocks = 0; - p->size = 0; - p->allocated = 0; -} - - -static SRes XzEncIndex_ReAlloc(CXzEncIndex *p, size_t newSize, ISzAllocPtr alloc) -{ - Byte *blocks = (Byte *)ISzAlloc_Alloc(alloc, newSize); - if (!blocks) - return SZ_ERROR_MEM; - if (p->size != 0) - memcpy(blocks, p->blocks, p->size); - if (p->blocks) - ISzAlloc_Free(alloc, p->blocks); - p->blocks = blocks; - p->allocated = newSize; - return SZ_OK; -} - - -static SRes XzEncIndex_PreAlloc(CXzEncIndex *p, UInt64 numBlocks, UInt64 unpackSize, UInt64 totalSize, ISzAllocPtr alloc) -{ - UInt64 pos; - { - Byte buf[32]; - unsigned pos2 = Xz_WriteVarInt(buf, totalSize); - pos2 += Xz_WriteVarInt(buf + pos2, unpackSize); - pos = numBlocks * pos2; - } - - if (pos <= p->allocated - p->size) - return SZ_OK; - { - UInt64 newSize64 = p->size + pos; - size_t newSize = (size_t)newSize64; - if (newSize != newSize64) - return SZ_ERROR_MEM; - return XzEncIndex_ReAlloc(p, newSize, alloc); - } -} - - -static SRes XzEncIndex_AddIndexRecord(CXzEncIndex *p, UInt64 unpackSize, UInt64 totalSize, ISzAllocPtr alloc) -{ - Byte buf[32]; - unsigned pos = Xz_WriteVarInt(buf, totalSize); - pos += Xz_WriteVarInt(buf + pos, unpackSize); - - if (pos > p->allocated - p->size) - { - size_t newSize = p->allocated * 2 + 16 * 2; - if (newSize < p->size + pos) - return SZ_ERROR_MEM; - RINOK(XzEncIndex_ReAlloc(p, newSize, alloc)); - } - memcpy(p->blocks + p->size, buf, pos); - p->size += pos; - p->numBlocks++; - return SZ_OK; -} - - -static SRes XzEncIndex_WriteFooter(const CXzEncIndex *p, CXzStreamFlags flags, ISeqOutStream *s) -{ - Byte buf[32]; - UInt64 globalPos; - UInt32 crc = CRC_INIT_VAL; - unsigned pos = 1 + Xz_WriteVarInt(buf + 1, p->numBlocks); - - globalPos = pos; - buf[0] = 0; - RINOK(WriteBytesUpdateCrc(s, buf, pos, &crc)); - RINOK(WriteBytesUpdateCrc(s, p->blocks, p->size, &crc)); - globalPos += p->size; - - pos = XZ_GET_PAD_SIZE(globalPos); - buf[1] = 0; - buf[2] = 0; - buf[3] = 0; - globalPos += pos; - - crc = CrcUpdate(crc, buf + 4 - pos, pos); - SetUi32(buf + 4, CRC_GET_DIGEST(crc)); - - SetUi32(buf + 8 + 4, (UInt32)(globalPos >> 2)); - buf[8 + 8] = (Byte)(flags >> 8); - buf[8 + 9] = (Byte)(flags & 0xFF); - SetUi32(buf + 8, CrcCalc(buf + 8 + 4, 6)); - buf[8 + 10] = XZ_FOOTER_SIG_0; - buf[8 + 11] = XZ_FOOTER_SIG_1; - - return WriteBytes(s, buf + 4 - pos, pos + 4 + 12); -} - - - -/* ---------- CSeqCheckInStream ---------- */ - -typedef struct -{ - ISeqInStream vt; - ISeqInStream *realStream; - const Byte *data; - UInt64 limit; - UInt64 processed; - int realStreamFinished; - CXzCheck check; -} CSeqCheckInStream; - -static void SeqCheckInStream_Init(CSeqCheckInStream *p, unsigned checkMode) -{ - p->limit = (UInt64)(Int64)-1; - p->processed = 0; - p->realStreamFinished = 0; - XzCheck_Init(&p->check, checkMode); -} - -static void SeqCheckInStream_GetDigest(CSeqCheckInStream *p, Byte *digest) -{ - XzCheck_Final(&p->check, digest); -} - -static SRes SeqCheckInStream_Read(const ISeqInStream *pp, void *data, size_t *size) -{ - CSeqCheckInStream *p = CONTAINER_FROM_VTBL(pp, CSeqCheckInStream, vt); - size_t size2 = *size; - SRes res = SZ_OK; - - if (p->limit != (UInt64)(Int64)-1) - { - UInt64 rem = p->limit - p->processed; - if (size2 > rem) - size2 = (size_t)rem; - } - if (size2 != 0) - { - if (p->realStream) - { - res = ISeqInStream_Read(p->realStream, data, &size2); - p->realStreamFinished = (size2 == 0) ? 1 : 0; - } - else - memcpy(data, p->data + (size_t)p->processed, size2); - XzCheck_Update(&p->check, data, size2); - p->processed += size2; - } - *size = size2; - return res; -} - - -/* ---------- CSeqSizeOutStream ---------- */ - -typedef struct -{ - ISeqOutStream vt; - ISeqOutStream *realStream; - Byte *outBuf; - size_t outBufLimit; - UInt64 processed; -} CSeqSizeOutStream; - -static size_t SeqSizeOutStream_Write(const ISeqOutStream *pp, const void *data, size_t size) -{ - CSeqSizeOutStream *p = CONTAINER_FROM_VTBL(pp, CSeqSizeOutStream, vt); - if (p->realStream) - size = ISeqOutStream_Write(p->realStream, data, size); - else - { - if (size > p->outBufLimit - (size_t)p->processed) - return 0; - memcpy(p->outBuf + (size_t)p->processed, data, size); - } - p->processed += size; - return size; -} - - -/* ---------- CSeqInFilter ---------- */ - -#define FILTER_BUF_SIZE (1 << 20) - -typedef struct -{ - ISeqInStream p; - ISeqInStream *realStream; - IStateCoder StateCoder; - Byte *buf; - size_t curPos; - size_t endPos; - int srcWasFinished; -} CSeqInFilter; - - -SRes BraState_SetFromMethod(IStateCoder *p, UInt64 id, int encodeMode, ISzAllocPtr alloc); - -static SRes SeqInFilter_Init(CSeqInFilter *p, const CXzFilter *props, ISzAllocPtr alloc) -{ - if (!p->buf) - { - p->buf = (Byte *)ISzAlloc_Alloc(alloc, FILTER_BUF_SIZE); - if (!p->buf) - return SZ_ERROR_MEM; - } - p->curPos = p->endPos = 0; - p->srcWasFinished = 0; - RINOK(BraState_SetFromMethod(&p->StateCoder, props->id, 1, alloc)); - RINOK(p->StateCoder.SetProps(p->StateCoder.p, props->props, props->propsSize, alloc)); - p->StateCoder.Init(p->StateCoder.p); - return SZ_OK; -} - - -static SRes SeqInFilter_Read(const ISeqInStream *pp, void *data, size_t *size) -{ - CSeqInFilter *p = CONTAINER_FROM_VTBL(pp, CSeqInFilter, p); - size_t sizeOriginal = *size; - if (sizeOriginal == 0) - return SZ_OK; - *size = 0; - - for (;;) - { - if (!p->srcWasFinished && p->curPos == p->endPos) - { - p->curPos = 0; - p->endPos = FILTER_BUF_SIZE; - RINOK(ISeqInStream_Read(p->realStream, p->buf, &p->endPos)); - if (p->endPos == 0) - p->srcWasFinished = 1; - } - { - SizeT srcLen = p->endPos - p->curPos; - ECoderStatus status; - SRes res; - *size = sizeOriginal; - res = p->StateCoder.Code2(p->StateCoder.p, - (Byte *)data, size, - p->buf + p->curPos, &srcLen, - p->srcWasFinished, CODER_FINISH_ANY, - &status); - p->curPos += srcLen; - if (*size != 0 || srcLen == 0 || res != SZ_OK) - return res; - } - } -} - -static void SeqInFilter_Construct(CSeqInFilter *p) -{ - p->buf = NULL; - p->StateCoder.p = NULL; - p->p.Read = SeqInFilter_Read; -} - -static void SeqInFilter_Free(CSeqInFilter *p, ISzAllocPtr alloc) -{ - if (p->StateCoder.p) - { - p->StateCoder.Free(p->StateCoder.p, alloc); - p->StateCoder.p = NULL; - } - if (p->buf) - { - ISzAlloc_Free(alloc, p->buf); - p->buf = NULL; - } -} - - -/* ---------- CSbEncInStream ---------- */ - -#ifdef USE_SUBBLOCK - -typedef struct -{ - ISeqInStream vt; - ISeqInStream *inStream; - CSbEnc enc; -} CSbEncInStream; - -static SRes SbEncInStream_Read(const ISeqInStream *pp, void *data, size_t *size) -{ - CSbEncInStream *p = CONTAINER_FROM_VTBL(pp, CSbEncInStream, vt); - size_t sizeOriginal = *size; - if (sizeOriginal == 0) - return SZ_OK; - - for (;;) - { - if (p->enc.needRead && !p->enc.readWasFinished) - { - size_t processed = p->enc.needReadSizeMax; - RINOK(p->inStream->Read(p->inStream, p->enc.buf + p->enc.readPos, &processed)); - p->enc.readPos += processed; - if (processed == 0) - { - p->enc.readWasFinished = True; - p->enc.isFinalFinished = True; - } - p->enc.needRead = False; - } - - *size = sizeOriginal; - RINOK(SbEnc_Read(&p->enc, data, size)); - if (*size != 0 || !p->enc.needRead) - return SZ_OK; - } -} - -void SbEncInStream_Construct(CSbEncInStream *p, ISzAllocPtr alloc) -{ - SbEnc_Construct(&p->enc, alloc); - p->vt.Read = SbEncInStream_Read; -} - -SRes SbEncInStream_Init(CSbEncInStream *p) -{ - return SbEnc_Init(&p->enc); -} - -void SbEncInStream_Free(CSbEncInStream *p) -{ - SbEnc_Free(&p->enc); -} - -#endif - - - -/* ---------- CXzProps ---------- */ - - -void XzFilterProps_Init(CXzFilterProps *p) -{ - p->id = 0; - p->delta = 0; - p->ip = 0; - p->ipDefined = False; -} - -void XzProps_Init(CXzProps *p) -{ - p->checkId = XZ_CHECK_CRC32; - p->blockSize = XZ_PROPS__BLOCK_SIZE__AUTO; - p->numBlockThreads_Reduced = -1; - p->numBlockThreads_Max = -1; - p->numTotalThreads = -1; - p->reduceSize = (UInt64)(Int64)-1; - p->forceWriteSizesInHeader = 0; - // p->forceWriteSizesInHeader = 1; - - XzFilterProps_Init(&p->filterProps); - Lzma2EncProps_Init(&p->lzma2Props); -} - - -static void XzEncProps_Normalize_Fixed(CXzProps *p) -{ - UInt64 fileSize; - int t1, t1n, t2, t2r, t3; - { - CLzma2EncProps tp = p->lzma2Props; - if (tp.numTotalThreads <= 0) - tp.numTotalThreads = p->numTotalThreads; - Lzma2EncProps_Normalize(&tp); - t1n = tp.numTotalThreads; - } - - t1 = p->lzma2Props.numTotalThreads; - t2 = p->numBlockThreads_Max; - t3 = p->numTotalThreads; - - if (t2 > MTCODER__THREADS_MAX) - t2 = MTCODER__THREADS_MAX; - - if (t3 <= 0) - { - if (t2 <= 0) - t2 = 1; - t3 = t1n * t2; - } - else if (t2 <= 0) - { - t2 = t3 / t1n; - if (t2 == 0) - { - t1 = 1; - t2 = t3; - } - if (t2 > MTCODER__THREADS_MAX) - t2 = MTCODER__THREADS_MAX; - } - else if (t1 <= 0) - { - t1 = t3 / t2; - if (t1 == 0) - t1 = 1; - } - else - t3 = t1n * t2; - - p->lzma2Props.numTotalThreads = t1; - - t2r = t2; - - fileSize = p->reduceSize; - - if ((p->blockSize < fileSize || fileSize == (UInt64)(Int64)-1)) - p->lzma2Props.lzmaProps.reduceSize = p->blockSize; - - Lzma2EncProps_Normalize(&p->lzma2Props); - - t1 = p->lzma2Props.numTotalThreads; - - { - if (t2 > 1 && fileSize != (UInt64)(Int64)-1) - { - UInt64 numBlocks = fileSize / p->blockSize; - if (numBlocks * p->blockSize != fileSize) - numBlocks++; - if (numBlocks < (unsigned)t2) - { - t2r = (int)numBlocks; - if (t2r == 0) - t2r = 1; - t3 = t1 * t2r; - } - } - } - - p->numBlockThreads_Max = t2; - p->numBlockThreads_Reduced = t2r; - p->numTotalThreads = t3; -} - - -static void XzProps_Normalize(CXzProps *p) -{ - /* we normalize xzProps properties, but we normalize only some of CXzProps::lzma2Props properties. - Lzma2Enc_SetProps() will normalize lzma2Props later. */ - - if (p->blockSize == XZ_PROPS__BLOCK_SIZE__SOLID) - { - p->lzma2Props.lzmaProps.reduceSize = p->reduceSize; - p->numBlockThreads_Reduced = 1; - p->numBlockThreads_Max = 1; - if (p->lzma2Props.numTotalThreads <= 0) - p->lzma2Props.numTotalThreads = p->numTotalThreads; - return; - } - else - { - CLzma2EncProps *lzma2 = &p->lzma2Props; - if (p->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO) - { - // xz-auto - p->lzma2Props.lzmaProps.reduceSize = p->reduceSize; - - if (lzma2->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID) - { - // if (xz-auto && lzma2-solid) - we use solid for both - p->blockSize = XZ_PROPS__BLOCK_SIZE__SOLID; - p->numBlockThreads_Reduced = 1; - p->numBlockThreads_Max = 1; - if (p->lzma2Props.numTotalThreads <= 0) - p->lzma2Props.numTotalThreads = p->numTotalThreads; - } - else - { - // if (xz-auto && (lzma2-auto || lzma2-fixed_) - // we calculate block size for lzma2 and use that block size for xz, lzma2 uses single-chunk per block - CLzma2EncProps tp = p->lzma2Props; - if (tp.numTotalThreads <= 0) - tp.numTotalThreads = p->numTotalThreads; - - Lzma2EncProps_Normalize(&tp); - - p->blockSize = tp.blockSize; // fixed or solid - p->numBlockThreads_Reduced = tp.numBlockThreads_Reduced; - p->numBlockThreads_Max = tp.numBlockThreads_Max; - if (lzma2->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO) - lzma2->blockSize = tp.blockSize; // fixed or solid, LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID - if (lzma2->lzmaProps.reduceSize > tp.blockSize && tp.blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID) - lzma2->lzmaProps.reduceSize = tp.blockSize; - lzma2->numBlockThreads_Reduced = 1; - lzma2->numBlockThreads_Max = 1; - return; - } - } - else - { - // xz-fixed - // we can use xz::reduceSize or xz::blockSize as base for lzmaProps::reduceSize - - p->lzma2Props.lzmaProps.reduceSize = p->reduceSize; - { - UInt64 r = p->reduceSize; - if (r > p->blockSize || r == (UInt64)(Int64)-1) - r = p->blockSize; - lzma2->lzmaProps.reduceSize = r; - } - if (lzma2->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO) - lzma2->blockSize = LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID; - else if (lzma2->blockSize > p->blockSize && lzma2->blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID) - lzma2->blockSize = p->blockSize; - - XzEncProps_Normalize_Fixed(p); - } - } -} - - -/* ---------- CLzma2WithFilters ---------- */ - -typedef struct -{ - CLzma2EncHandle lzma2; - CSeqInFilter filter; - - #ifdef USE_SUBBLOCK - CSbEncInStream sb; - #endif -} CLzma2WithFilters; - - -static void Lzma2WithFilters_Construct(CLzma2WithFilters *p) -{ - p->lzma2 = NULL; - SeqInFilter_Construct(&p->filter); - - #ifdef USE_SUBBLOCK - SbEncInStream_Construct(&p->sb, alloc); - #endif -} - - -static SRes Lzma2WithFilters_Create(CLzma2WithFilters *p, ISzAllocPtr alloc, ISzAllocPtr bigAlloc) -{ - if (!p->lzma2) - { - p->lzma2 = Lzma2Enc_Create(alloc, bigAlloc); - if (!p->lzma2) - return SZ_ERROR_MEM; - } - return SZ_OK; -} - - -static void Lzma2WithFilters_Free(CLzma2WithFilters *p, ISzAllocPtr alloc) -{ - #ifdef USE_SUBBLOCK - SbEncInStream_Free(&p->sb); - #endif - - SeqInFilter_Free(&p->filter, alloc); - if (p->lzma2) - { - Lzma2Enc_Destroy(p->lzma2); - p->lzma2 = NULL; - } -} - - -typedef struct -{ - UInt64 unpackSize; - UInt64 totalSize; - size_t headerSize; -} CXzEncBlockInfo; - - -static SRes Xz_CompressBlock( - CLzma2WithFilters *lzmaf, - - ISeqOutStream *outStream, - Byte *outBufHeader, - Byte *outBufData, size_t outBufDataLimit, - - ISeqInStream *inStream, - // UInt64 expectedSize, - const Byte *inBuf, // used if (!inStream) - size_t inBufSize, // used if (!inStream), it's block size, props->blockSize is ignored - - const CXzProps *props, - ICompressProgress *progress, - int *inStreamFinished, /* only for inStream version */ - CXzEncBlockInfo *blockSizes, - ISzAllocPtr alloc, - ISzAllocPtr allocBig) -{ - CSeqCheckInStream checkInStream; - CSeqSizeOutStream seqSizeOutStream; - CXzBlock block; - unsigned filterIndex = 0; - CXzFilter *filter = NULL; - const CXzFilterProps *fp = &props->filterProps; - if (fp->id == 0) - fp = NULL; - - *inStreamFinished = False; - - RINOK(Lzma2WithFilters_Create(lzmaf, alloc, allocBig)); - - RINOK(Lzma2Enc_SetProps(lzmaf->lzma2, &props->lzma2Props)); - - XzBlock_ClearFlags(&block); - XzBlock_SetNumFilters(&block, 1 + (fp ? 1 : 0)); - - if (fp) - { - filter = &block.filters[filterIndex++]; - filter->id = fp->id; - filter->propsSize = 0; - - if (fp->id == XZ_ID_Delta) - { - filter->props[0] = (Byte)(fp->delta - 1); - filter->propsSize = 1; - } - else if (fp->ipDefined) - { - Byte *ptr = filter->props; - SetUi32(ptr, fp->ip); - filter->propsSize = 4; - } - } - - { - CXzFilter *f = &block.filters[filterIndex++]; - f->id = XZ_ID_LZMA2; - f->propsSize = 1; - f->props[0] = Lzma2Enc_WriteProperties(lzmaf->lzma2); - } - - seqSizeOutStream.vt.Write = SeqSizeOutStream_Write; - seqSizeOutStream.realStream = outStream; - seqSizeOutStream.outBuf = outBufData; - seqSizeOutStream.outBufLimit = outBufDataLimit; - seqSizeOutStream.processed = 0; - - /* - if (expectedSize != (UInt64)(Int64)-1) - { - block.unpackSize = expectedSize; - if (props->blockSize != (UInt64)(Int64)-1) - if (expectedSize > props->blockSize) - block.unpackSize = props->blockSize; - XzBlock_SetHasUnpackSize(&block); - } - */ - - if (outStream) - { - RINOK(XzBlock_WriteHeader(&block, &seqSizeOutStream.vt)); - } - - checkInStream.vt.Read = SeqCheckInStream_Read; - SeqCheckInStream_Init(&checkInStream, props->checkId); - - checkInStream.realStream = inStream; - checkInStream.data = inBuf; - checkInStream.limit = props->blockSize; - if (!inStream) - checkInStream.limit = inBufSize; - - if (fp) - { - #ifdef USE_SUBBLOCK - if (fp->id == XZ_ID_Subblock) - { - lzmaf->sb.inStream = &checkInStream.vt; - RINOK(SbEncInStream_Init(&lzmaf->sb)); - } - else - #endif - { - lzmaf->filter.realStream = &checkInStream.vt; - RINOK(SeqInFilter_Init(&lzmaf->filter, filter, alloc)); - } - } - - { - SRes res; - Byte *outBuf = NULL; - size_t outSize = 0; - BoolInt useStream = (fp || inStream); - // useStream = True; - - if (!useStream) - { - XzCheck_Update(&checkInStream.check, inBuf, inBufSize); - checkInStream.processed = inBufSize; - } - - if (!outStream) - { - outBuf = seqSizeOutStream.outBuf; // + (size_t)seqSizeOutStream.processed; - outSize = seqSizeOutStream.outBufLimit; // - (size_t)seqSizeOutStream.processed; - } - - res = Lzma2Enc_Encode2(lzmaf->lzma2, - outBuf ? NULL : &seqSizeOutStream.vt, - outBuf, - outBuf ? &outSize : NULL, - - useStream ? - (fp ? - ( - #ifdef USE_SUBBLOCK - (fp->id == XZ_ID_Subblock) ? &lzmaf->sb.vt: - #endif - &lzmaf->filter.p) : - &checkInStream.vt) : NULL, - - useStream ? NULL : inBuf, - useStream ? 0 : inBufSize, - - progress); - - if (outBuf) - seqSizeOutStream.processed += outSize; - - RINOK(res); - blockSizes->unpackSize = checkInStream.processed; - } - { - Byte buf[4 + 64]; - unsigned padSize = XZ_GET_PAD_SIZE(seqSizeOutStream.processed); - UInt64 packSize = seqSizeOutStream.processed; - - buf[0] = 0; - buf[1] = 0; - buf[2] = 0; - buf[3] = 0; - - SeqCheckInStream_GetDigest(&checkInStream, buf + 4); - RINOK(WriteBytes(&seqSizeOutStream.vt, buf + (4 - padSize), padSize + XzFlags_GetCheckSize((CXzStreamFlags)props->checkId))); - - blockSizes->totalSize = seqSizeOutStream.processed - padSize; - - if (!outStream) - { - seqSizeOutStream.outBuf = outBufHeader; - seqSizeOutStream.outBufLimit = XZ_BLOCK_HEADER_SIZE_MAX; - seqSizeOutStream.processed = 0; - - block.unpackSize = blockSizes->unpackSize; - XzBlock_SetHasUnpackSize(&block); - - block.packSize = packSize; - XzBlock_SetHasPackSize(&block); - - RINOK(XzBlock_WriteHeader(&block, &seqSizeOutStream.vt)); - - blockSizes->headerSize = (size_t)seqSizeOutStream.processed; - blockSizes->totalSize += seqSizeOutStream.processed; - } - } - - if (inStream) - *inStreamFinished = checkInStream.realStreamFinished; - else - { - *inStreamFinished = False; - if (checkInStream.processed != inBufSize) - return SZ_ERROR_FAIL; - } - - return SZ_OK; -} - - - -typedef struct -{ - ICompressProgress vt; - ICompressProgress *progress; - UInt64 inOffset; - UInt64 outOffset; -} CCompressProgress_XzEncOffset; - - -static SRes CompressProgress_XzEncOffset_Progress(const ICompressProgress *pp, UInt64 inSize, UInt64 outSize) -{ - const CCompressProgress_XzEncOffset *p = CONTAINER_FROM_VTBL(pp, CCompressProgress_XzEncOffset, vt); - inSize += p->inOffset; - outSize += p->outOffset; - return ICompressProgress_Progress(p->progress, inSize, outSize); -} - - - - -typedef struct -{ - ISzAllocPtr alloc; - ISzAllocPtr allocBig; - - CXzProps xzProps; - UInt64 expectedDataSize; - - CXzEncIndex xzIndex; - - CLzma2WithFilters lzmaf_Items[MTCODER__THREADS_MAX]; - - size_t outBufSize; /* size of allocated outBufs[i] */ - Byte *outBufs[MTCODER__BLOCKS_MAX]; - - #ifndef _7ZIP_ST - unsigned checkType; - ISeqOutStream *outStream; - BoolInt mtCoder_WasConstructed; - CMtCoder mtCoder; - CXzEncBlockInfo EncBlocks[MTCODER__BLOCKS_MAX]; - #endif - -} CXzEnc; - - -static void XzEnc_Construct(CXzEnc *p) -{ - unsigned i; - - XzEncIndex_Construct(&p->xzIndex); - - for (i = 0; i < MTCODER__THREADS_MAX; i++) - Lzma2WithFilters_Construct(&p->lzmaf_Items[i]); - - #ifndef _7ZIP_ST - p->mtCoder_WasConstructed = False; - { - for (i = 0; i < MTCODER__BLOCKS_MAX; i++) - p->outBufs[i] = NULL; - p->outBufSize = 0; - } - #endif -} - - -static void XzEnc_FreeOutBufs(CXzEnc *p) -{ - unsigned i; - for (i = 0; i < MTCODER__BLOCKS_MAX; i++) - if (p->outBufs[i]) - { - ISzAlloc_Free(p->alloc, p->outBufs[i]); - p->outBufs[i] = NULL; - } - p->outBufSize = 0; -} - - -static void XzEnc_Free(CXzEnc *p, ISzAllocPtr alloc) -{ - unsigned i; - - XzEncIndex_Free(&p->xzIndex, alloc); - - for (i = 0; i < MTCODER__THREADS_MAX; i++) - Lzma2WithFilters_Free(&p->lzmaf_Items[i], alloc); - - #ifndef _7ZIP_ST - if (p->mtCoder_WasConstructed) - { - MtCoder_Destruct(&p->mtCoder); - p->mtCoder_WasConstructed = False; - } - XzEnc_FreeOutBufs(p); - #endif -} - - -CXzEncHandle XzEnc_Create(ISzAllocPtr alloc, ISzAllocPtr allocBig) -{ - CXzEnc *p = (CXzEnc *)ISzAlloc_Alloc(alloc, sizeof(CXzEnc)); - if (!p) - return NULL; - XzEnc_Construct(p); - XzProps_Init(&p->xzProps); - XzProps_Normalize(&p->xzProps); - p->expectedDataSize = (UInt64)(Int64)-1; - p->alloc = alloc; - p->allocBig = allocBig; - return p; -} - - -void XzEnc_Destroy(CXzEncHandle pp) -{ - CXzEnc *p = (CXzEnc *)pp; - XzEnc_Free(p, p->alloc); - ISzAlloc_Free(p->alloc, p); -} - - -SRes XzEnc_SetProps(CXzEncHandle pp, const CXzProps *props) -{ - CXzEnc *p = (CXzEnc *)pp; - p->xzProps = *props; - XzProps_Normalize(&p->xzProps); - return SZ_OK; -} - - -void XzEnc_SetDataSize(CXzEncHandle pp, UInt64 expectedDataSiize) -{ - CXzEnc *p = (CXzEnc *)pp; - p->expectedDataSize = expectedDataSiize; -} - - - - -#ifndef _7ZIP_ST - -static SRes XzEnc_MtCallback_Code(void *pp, unsigned coderIndex, unsigned outBufIndex, - const Byte *src, size_t srcSize, int finished) -{ - CXzEnc *me = (CXzEnc *)pp; - SRes res; - CMtProgressThunk progressThunk; - - Byte *dest = me->outBufs[outBufIndex]; - - UNUSED_VAR(finished) - - { - CXzEncBlockInfo *bInfo = &me->EncBlocks[outBufIndex]; - bInfo->totalSize = 0; - bInfo->unpackSize = 0; - bInfo->headerSize = 0; - } - - if (!dest) - { - dest = (Byte *)ISzAlloc_Alloc(me->alloc, me->outBufSize); - if (!dest) - return SZ_ERROR_MEM; - me->outBufs[outBufIndex] = dest; - } - - MtProgressThunk_CreateVTable(&progressThunk); - progressThunk.mtProgress = &me->mtCoder.mtProgress; - MtProgressThunk_Init(&progressThunk); - - { - CXzEncBlockInfo blockSizes; - int inStreamFinished; - - res = Xz_CompressBlock( - &me->lzmaf_Items[coderIndex], - - NULL, - dest, - dest + XZ_BLOCK_HEADER_SIZE_MAX, me->outBufSize - XZ_BLOCK_HEADER_SIZE_MAX, - - NULL, - // srcSize, // expectedSize - src, srcSize, - - &me->xzProps, - &progressThunk.vt, - &inStreamFinished, - &blockSizes, - me->alloc, - me->allocBig); - - if (res == SZ_OK) - me->EncBlocks[outBufIndex] = blockSizes; - - return res; - } -} - - -static SRes XzEnc_MtCallback_Write(void *pp, unsigned outBufIndex) -{ - CXzEnc *me = (CXzEnc *)pp; - - const CXzEncBlockInfo *bInfo = &me->EncBlocks[outBufIndex]; - const Byte *data = me->outBufs[outBufIndex]; - - RINOK(WriteBytes(me->outStream, data, bInfo->headerSize)); - - { - UInt64 totalPackFull = bInfo->totalSize + XZ_GET_PAD_SIZE(bInfo->totalSize); - RINOK(WriteBytes(me->outStream, data + XZ_BLOCK_HEADER_SIZE_MAX, (size_t)totalPackFull - bInfo->headerSize)); - } - - return XzEncIndex_AddIndexRecord(&me->xzIndex, bInfo->unpackSize, bInfo->totalSize, me->alloc); -} - -#endif - - - -SRes XzEnc_Encode(CXzEncHandle pp, ISeqOutStream *outStream, ISeqInStream *inStream, ICompressProgress *progress) -{ - CXzEnc *p = (CXzEnc *)pp; - - const CXzProps *props = &p->xzProps; - - XzEncIndex_Init(&p->xzIndex); - { - UInt64 numBlocks = 1; - UInt64 blockSize = props->blockSize; - - if (blockSize != XZ_PROPS__BLOCK_SIZE__SOLID - && props->reduceSize != (UInt64)(Int64)-1) - { - numBlocks = props->reduceSize / blockSize; - if (numBlocks * blockSize != props->reduceSize) - numBlocks++; - } - else - blockSize = (UInt64)1 << 62; - - RINOK(XzEncIndex_PreAlloc(&p->xzIndex, numBlocks, blockSize, XZ_GET_ESTIMATED_BLOCK_TOTAL_PACK_SIZE(blockSize), p->alloc)); - } - - RINOK(Xz_WriteHeader((CXzStreamFlags)props->checkId, outStream)); - - - #ifndef _7ZIP_ST - if (props->numBlockThreads_Reduced > 1) - { - IMtCoderCallback2 vt; - - if (!p->mtCoder_WasConstructed) - { - p->mtCoder_WasConstructed = True; - MtCoder_Construct(&p->mtCoder); - } - - vt.Code = XzEnc_MtCallback_Code; - vt.Write = XzEnc_MtCallback_Write; - - p->checkType = props->checkId; - p->xzProps = *props; - - p->outStream = outStream; - - p->mtCoder.allocBig = p->allocBig; - p->mtCoder.progress = progress; - p->mtCoder.inStream = inStream; - p->mtCoder.inData = NULL; - p->mtCoder.inDataSize = 0; - p->mtCoder.mtCallback = &vt; - p->mtCoder.mtCallbackObject = p; - - if ( props->blockSize == XZ_PROPS__BLOCK_SIZE__SOLID - || props->blockSize == XZ_PROPS__BLOCK_SIZE__AUTO) - return SZ_ERROR_FAIL; - - p->mtCoder.blockSize = (size_t)props->blockSize; - if (p->mtCoder.blockSize != props->blockSize) - return SZ_ERROR_PARAM; /* SZ_ERROR_MEM */ - - { - size_t destBlockSize = XZ_BLOCK_HEADER_SIZE_MAX + XZ_GET_MAX_BLOCK_PACK_SIZE(p->mtCoder.blockSize); - if (destBlockSize < p->mtCoder.blockSize) - return SZ_ERROR_PARAM; - if (p->outBufSize != destBlockSize) - XzEnc_FreeOutBufs(p); - p->outBufSize = destBlockSize; - } - - p->mtCoder.numThreadsMax = (unsigned)props->numBlockThreads_Max; - p->mtCoder.expectedDataSize = p->expectedDataSize; - - RINOK(MtCoder_Code(&p->mtCoder)); - } - else - #endif - { - int writeStartSizes; - CCompressProgress_XzEncOffset progress2; - Byte *bufData = NULL; - size_t bufSize = 0; - - progress2.vt.Progress = CompressProgress_XzEncOffset_Progress; - progress2.inOffset = 0; - progress2.outOffset = 0; - progress2.progress = progress; - - writeStartSizes = 0; - - if (props->blockSize != XZ_PROPS__BLOCK_SIZE__SOLID) - { - writeStartSizes = (props->forceWriteSizesInHeader > 0); - - if (writeStartSizes) - { - size_t t2; - size_t t = (size_t)props->blockSize; - if (t != props->blockSize) - return SZ_ERROR_PARAM; - t = XZ_GET_MAX_BLOCK_PACK_SIZE(t); - if (t < props->blockSize) - return SZ_ERROR_PARAM; - t2 = XZ_BLOCK_HEADER_SIZE_MAX + t; - if (!p->outBufs[0] || t2 != p->outBufSize) - { - XzEnc_FreeOutBufs(p); - p->outBufs[0] = (Byte *)ISzAlloc_Alloc(p->alloc, t2); - if (!p->outBufs[0]) - return SZ_ERROR_MEM; - p->outBufSize = t2; - } - bufData = p->outBufs[0] + XZ_BLOCK_HEADER_SIZE_MAX; - bufSize = t; - } - } - - for (;;) - { - CXzEncBlockInfo blockSizes; - int inStreamFinished; - - /* - UInt64 rem = (UInt64)(Int64)-1; - if (props->reduceSize != (UInt64)(Int64)-1 - && props->reduceSize >= progress2.inOffset) - rem = props->reduceSize - progress2.inOffset; - */ - - blockSizes.headerSize = 0; // for GCC - - RINOK(Xz_CompressBlock( - &p->lzmaf_Items[0], - - writeStartSizes ? NULL : outStream, - writeStartSizes ? p->outBufs[0] : NULL, - bufData, bufSize, - - inStream, - // rem, - NULL, 0, - - props, - progress ? &progress2.vt : NULL, - &inStreamFinished, - &blockSizes, - p->alloc, - p->allocBig)); - - { - UInt64 totalPackFull = blockSizes.totalSize + XZ_GET_PAD_SIZE(blockSizes.totalSize); - - if (writeStartSizes) - { - RINOK(WriteBytes(outStream, p->outBufs[0], blockSizes.headerSize)); - RINOK(WriteBytes(outStream, bufData, (size_t)totalPackFull - blockSizes.headerSize)); - } - - RINOK(XzEncIndex_AddIndexRecord(&p->xzIndex, blockSizes.unpackSize, blockSizes.totalSize, p->alloc)); - - progress2.inOffset += blockSizes.unpackSize; - progress2.outOffset += totalPackFull; - } - - if (inStreamFinished) - break; - } - } - - return XzEncIndex_WriteFooter(&p->xzIndex, (CXzStreamFlags)props->checkId, outStream); -} - - -#include "Alloc.h" - -SRes Xz_Encode(ISeqOutStream *outStream, ISeqInStream *inStream, - const CXzProps *props, ICompressProgress *progress) -{ - SRes res; - CXzEncHandle xz = XzEnc_Create(&g_Alloc, &g_BigAlloc); - if (!xz) - return SZ_ERROR_MEM; - res = XzEnc_SetProps(xz, props); - if (res == SZ_OK) - res = XzEnc_Encode(xz, outStream, inStream, progress); - XzEnc_Destroy(xz); - return res; -} - - -SRes Xz_EncodeEmpty(ISeqOutStream *outStream) -{ - SRes res; - CXzEncIndex xzIndex; - XzEncIndex_Construct(&xzIndex); - res = Xz_WriteHeader((CXzStreamFlags)0, outStream); - if (res == SZ_OK) - res = XzEncIndex_WriteFooter(&xzIndex, (CXzStreamFlags)0, outStream); - XzEncIndex_Free(&xzIndex, NULL); // g_Alloc - return res; -} +/* XzEnc.c -- Xz Encode +2021-04-01 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include +#include + +#include "7zCrc.h" +#include "Bra.h" +#include "CpuArch.h" + +#ifdef USE_SUBBLOCK +#include "Bcj3Enc.c" +#include "SbFind.c" +#include "SbEnc.c" +#endif + +#include "XzEnc.h" + +// #define _7ZIP_ST + +#ifndef _7ZIP_ST +#include "MtCoder.h" +#else +#define MTCODER__THREADS_MAX 1 +#define MTCODER__BLOCKS_MAX 1 +#endif + +#define XZ_GET_PAD_SIZE(dataSize) ((4 - ((unsigned)(dataSize) & 3)) & 3) + +/* max pack size for LZMA2 block + check-64bytrs: */ +#define XZ_GET_MAX_BLOCK_PACK_SIZE(unpackSize) ((unpackSize) + ((unpackSize) >> 10) + 16 + 64) + +#define XZ_GET_ESTIMATED_BLOCK_TOTAL_PACK_SIZE(unpackSize) (XZ_BLOCK_HEADER_SIZE_MAX + XZ_GET_MAX_BLOCK_PACK_SIZE(unpackSize)) + + +#define XzBlock_ClearFlags(p) (p)->flags = 0; +#define XzBlock_SetNumFilters(p, n) (p)->flags = (Byte)((p)->flags | ((n) - 1)); +#define XzBlock_SetHasPackSize(p) (p)->flags |= XZ_BF_PACK_SIZE; +#define XzBlock_SetHasUnpackSize(p) (p)->flags |= XZ_BF_UNPACK_SIZE; + + +static SRes WriteBytes(ISeqOutStream *s, const void *buf, size_t size) +{ + return (ISeqOutStream_Write(s, buf, size) == size) ? SZ_OK : SZ_ERROR_WRITE; +} + +static SRes WriteBytesUpdateCrc(ISeqOutStream *s, const void *buf, size_t size, UInt32 *crc) +{ + *crc = CrcUpdate(*crc, buf, size); + return WriteBytes(s, buf, size); +} + + +static SRes Xz_WriteHeader(CXzStreamFlags f, ISeqOutStream *s) +{ + UInt32 crc; + Byte header[XZ_STREAM_HEADER_SIZE]; + memcpy(header, XZ_SIG, XZ_SIG_SIZE); + header[XZ_SIG_SIZE] = (Byte)(f >> 8); + header[XZ_SIG_SIZE + 1] = (Byte)(f & 0xFF); + crc = CrcCalc(header + XZ_SIG_SIZE, XZ_STREAM_FLAGS_SIZE); + SetUi32(header + XZ_SIG_SIZE + XZ_STREAM_FLAGS_SIZE, crc); + return WriteBytes(s, header, XZ_STREAM_HEADER_SIZE); +} + + +static SRes XzBlock_WriteHeader(const CXzBlock *p, ISeqOutStream *s) +{ + Byte header[XZ_BLOCK_HEADER_SIZE_MAX]; + + unsigned pos = 1; + unsigned numFilters, i; + header[pos++] = p->flags; + + if (XzBlock_HasPackSize(p)) pos += Xz_WriteVarInt(header + pos, p->packSize); + if (XzBlock_HasUnpackSize(p)) pos += Xz_WriteVarInt(header + pos, p->unpackSize); + numFilters = XzBlock_GetNumFilters(p); + + for (i = 0; i < numFilters; i++) + { + const CXzFilter *f = &p->filters[i]; + pos += Xz_WriteVarInt(header + pos, f->id); + pos += Xz_WriteVarInt(header + pos, f->propsSize); + memcpy(header + pos, f->props, f->propsSize); + pos += f->propsSize; + } + + while ((pos & 3) != 0) + header[pos++] = 0; + + header[0] = (Byte)(pos >> 2); + SetUi32(header + pos, CrcCalc(header, pos)); + return WriteBytes(s, header, pos + 4); +} + + + + +typedef struct +{ + size_t numBlocks; + size_t size; + size_t allocated; + Byte *blocks; +} CXzEncIndex; + + +static void XzEncIndex_Construct(CXzEncIndex *p) +{ + p->numBlocks = 0; + p->size = 0; + p->allocated = 0; + p->blocks = NULL; +} + +static void XzEncIndex_Init(CXzEncIndex *p) +{ + p->numBlocks = 0; + p->size = 0; +} + +static void XzEncIndex_Free(CXzEncIndex *p, ISzAllocPtr alloc) +{ + if (p->blocks) + { + ISzAlloc_Free(alloc, p->blocks); + p->blocks = NULL; + } + p->numBlocks = 0; + p->size = 0; + p->allocated = 0; +} + + +static SRes XzEncIndex_ReAlloc(CXzEncIndex *p, size_t newSize, ISzAllocPtr alloc) +{ + Byte *blocks = (Byte *)ISzAlloc_Alloc(alloc, newSize); + if (!blocks) + return SZ_ERROR_MEM; + if (p->size != 0) + memcpy(blocks, p->blocks, p->size); + if (p->blocks) + ISzAlloc_Free(alloc, p->blocks); + p->blocks = blocks; + p->allocated = newSize; + return SZ_OK; +} + + +static SRes XzEncIndex_PreAlloc(CXzEncIndex *p, UInt64 numBlocks, UInt64 unpackSize, UInt64 totalSize, ISzAllocPtr alloc) +{ + UInt64 pos; + { + Byte buf[32]; + unsigned pos2 = Xz_WriteVarInt(buf, totalSize); + pos2 += Xz_WriteVarInt(buf + pos2, unpackSize); + pos = numBlocks * pos2; + } + + if (pos <= p->allocated - p->size) + return SZ_OK; + { + UInt64 newSize64 = p->size + pos; + size_t newSize = (size_t)newSize64; + if (newSize != newSize64) + return SZ_ERROR_MEM; + return XzEncIndex_ReAlloc(p, newSize, alloc); + } +} + + +static SRes XzEncIndex_AddIndexRecord(CXzEncIndex *p, UInt64 unpackSize, UInt64 totalSize, ISzAllocPtr alloc) +{ + Byte buf[32]; + unsigned pos = Xz_WriteVarInt(buf, totalSize); + pos += Xz_WriteVarInt(buf + pos, unpackSize); + + if (pos > p->allocated - p->size) + { + size_t newSize = p->allocated * 2 + 16 * 2; + if (newSize < p->size + pos) + return SZ_ERROR_MEM; + RINOK(XzEncIndex_ReAlloc(p, newSize, alloc)); + } + memcpy(p->blocks + p->size, buf, pos); + p->size += pos; + p->numBlocks++; + return SZ_OK; +} + + +static SRes XzEncIndex_WriteFooter(const CXzEncIndex *p, CXzStreamFlags flags, ISeqOutStream *s) +{ + Byte buf[32]; + UInt64 globalPos; + UInt32 crc = CRC_INIT_VAL; + unsigned pos = 1 + Xz_WriteVarInt(buf + 1, p->numBlocks); + + globalPos = pos; + buf[0] = 0; + RINOK(WriteBytesUpdateCrc(s, buf, pos, &crc)); + RINOK(WriteBytesUpdateCrc(s, p->blocks, p->size, &crc)); + globalPos += p->size; + + pos = XZ_GET_PAD_SIZE(globalPos); + buf[1] = 0; + buf[2] = 0; + buf[3] = 0; + globalPos += pos; + + crc = CrcUpdate(crc, buf + 4 - pos, pos); + SetUi32(buf + 4, CRC_GET_DIGEST(crc)); + + SetUi32(buf + 8 + 4, (UInt32)(globalPos >> 2)); + buf[8 + 8] = (Byte)(flags >> 8); + buf[8 + 9] = (Byte)(flags & 0xFF); + SetUi32(buf + 8, CrcCalc(buf + 8 + 4, 6)); + buf[8 + 10] = XZ_FOOTER_SIG_0; + buf[8 + 11] = XZ_FOOTER_SIG_1; + + return WriteBytes(s, buf + 4 - pos, pos + 4 + 12); +} + + + +/* ---------- CSeqCheckInStream ---------- */ + +typedef struct +{ + ISeqInStream vt; + ISeqInStream *realStream; + const Byte *data; + UInt64 limit; + UInt64 processed; + int realStreamFinished; + CXzCheck check; +} CSeqCheckInStream; + +static void SeqCheckInStream_Init(CSeqCheckInStream *p, unsigned checkMode) +{ + p->limit = (UInt64)(Int64)-1; + p->processed = 0; + p->realStreamFinished = 0; + XzCheck_Init(&p->check, checkMode); +} + +static void SeqCheckInStream_GetDigest(CSeqCheckInStream *p, Byte *digest) +{ + XzCheck_Final(&p->check, digest); +} + +static SRes SeqCheckInStream_Read(const ISeqInStream *pp, void *data, size_t *size) +{ + CSeqCheckInStream *p = CONTAINER_FROM_VTBL(pp, CSeqCheckInStream, vt); + size_t size2 = *size; + SRes res = SZ_OK; + + if (p->limit != (UInt64)(Int64)-1) + { + UInt64 rem = p->limit - p->processed; + if (size2 > rem) + size2 = (size_t)rem; + } + if (size2 != 0) + { + if (p->realStream) + { + res = ISeqInStream_Read(p->realStream, data, &size2); + p->realStreamFinished = (size2 == 0) ? 1 : 0; + } + else + memcpy(data, p->data + (size_t)p->processed, size2); + XzCheck_Update(&p->check, data, size2); + p->processed += size2; + } + *size = size2; + return res; +} + + +/* ---------- CSeqSizeOutStream ---------- */ + +typedef struct +{ + ISeqOutStream vt; + ISeqOutStream *realStream; + Byte *outBuf; + size_t outBufLimit; + UInt64 processed; +} CSeqSizeOutStream; + +static size_t SeqSizeOutStream_Write(const ISeqOutStream *pp, const void *data, size_t size) +{ + CSeqSizeOutStream *p = CONTAINER_FROM_VTBL(pp, CSeqSizeOutStream, vt); + if (p->realStream) + size = ISeqOutStream_Write(p->realStream, data, size); + else + { + if (size > p->outBufLimit - (size_t)p->processed) + return 0; + memcpy(p->outBuf + (size_t)p->processed, data, size); + } + p->processed += size; + return size; +} + + +/* ---------- CSeqInFilter ---------- */ + +#define FILTER_BUF_SIZE (1 << 20) + +typedef struct +{ + ISeqInStream p; + ISeqInStream *realStream; + IStateCoder StateCoder; + Byte *buf; + size_t curPos; + size_t endPos; + int srcWasFinished; +} CSeqInFilter; + + +SRes BraState_SetFromMethod(IStateCoder *p, UInt64 id, int encodeMode, ISzAllocPtr alloc); + +static SRes SeqInFilter_Init(CSeqInFilter *p, const CXzFilter *props, ISzAllocPtr alloc) +{ + if (!p->buf) + { + p->buf = (Byte *)ISzAlloc_Alloc(alloc, FILTER_BUF_SIZE); + if (!p->buf) + return SZ_ERROR_MEM; + } + p->curPos = p->endPos = 0; + p->srcWasFinished = 0; + RINOK(BraState_SetFromMethod(&p->StateCoder, props->id, 1, alloc)); + RINOK(p->StateCoder.SetProps(p->StateCoder.p, props->props, props->propsSize, alloc)); + p->StateCoder.Init(p->StateCoder.p); + return SZ_OK; +} + + +static SRes SeqInFilter_Read(const ISeqInStream *pp, void *data, size_t *size) +{ + CSeqInFilter *p = CONTAINER_FROM_VTBL(pp, CSeqInFilter, p); + size_t sizeOriginal = *size; + if (sizeOriginal == 0) + return SZ_OK; + *size = 0; + + for (;;) + { + if (!p->srcWasFinished && p->curPos == p->endPos) + { + p->curPos = 0; + p->endPos = FILTER_BUF_SIZE; + RINOK(ISeqInStream_Read(p->realStream, p->buf, &p->endPos)); + if (p->endPos == 0) + p->srcWasFinished = 1; + } + { + SizeT srcLen = p->endPos - p->curPos; + ECoderStatus status; + SRes res; + *size = sizeOriginal; + res = p->StateCoder.Code2(p->StateCoder.p, + (Byte *)data, size, + p->buf + p->curPos, &srcLen, + p->srcWasFinished, CODER_FINISH_ANY, + &status); + p->curPos += srcLen; + if (*size != 0 || srcLen == 0 || res != SZ_OK) + return res; + } + } +} + +static void SeqInFilter_Construct(CSeqInFilter *p) +{ + p->buf = NULL; + p->StateCoder.p = NULL; + p->p.Read = SeqInFilter_Read; +} + +static void SeqInFilter_Free(CSeqInFilter *p, ISzAllocPtr alloc) +{ + if (p->StateCoder.p) + { + p->StateCoder.Free(p->StateCoder.p, alloc); + p->StateCoder.p = NULL; + } + if (p->buf) + { + ISzAlloc_Free(alloc, p->buf); + p->buf = NULL; + } +} + + +/* ---------- CSbEncInStream ---------- */ + +#ifdef USE_SUBBLOCK + +typedef struct +{ + ISeqInStream vt; + ISeqInStream *inStream; + CSbEnc enc; +} CSbEncInStream; + +static SRes SbEncInStream_Read(const ISeqInStream *pp, void *data, size_t *size) +{ + CSbEncInStream *p = CONTAINER_FROM_VTBL(pp, CSbEncInStream, vt); + size_t sizeOriginal = *size; + if (sizeOriginal == 0) + return SZ_OK; + + for (;;) + { + if (p->enc.needRead && !p->enc.readWasFinished) + { + size_t processed = p->enc.needReadSizeMax; + RINOK(p->inStream->Read(p->inStream, p->enc.buf + p->enc.readPos, &processed)); + p->enc.readPos += processed; + if (processed == 0) + { + p->enc.readWasFinished = True; + p->enc.isFinalFinished = True; + } + p->enc.needRead = False; + } + + *size = sizeOriginal; + RINOK(SbEnc_Read(&p->enc, data, size)); + if (*size != 0 || !p->enc.needRead) + return SZ_OK; + } +} + +void SbEncInStream_Construct(CSbEncInStream *p, ISzAllocPtr alloc) +{ + SbEnc_Construct(&p->enc, alloc); + p->vt.Read = SbEncInStream_Read; +} + +SRes SbEncInStream_Init(CSbEncInStream *p) +{ + return SbEnc_Init(&p->enc); +} + +void SbEncInStream_Free(CSbEncInStream *p) +{ + SbEnc_Free(&p->enc); +} + +#endif + + + +/* ---------- CXzProps ---------- */ + + +void XzFilterProps_Init(CXzFilterProps *p) +{ + p->id = 0; + p->delta = 0; + p->ip = 0; + p->ipDefined = False; +} + +void XzProps_Init(CXzProps *p) +{ + p->checkId = XZ_CHECK_CRC32; + p->blockSize = XZ_PROPS__BLOCK_SIZE__AUTO; + p->numBlockThreads_Reduced = -1; + p->numBlockThreads_Max = -1; + p->numTotalThreads = -1; + p->reduceSize = (UInt64)(Int64)-1; + p->forceWriteSizesInHeader = 0; + // p->forceWriteSizesInHeader = 1; + + XzFilterProps_Init(&p->filterProps); + Lzma2EncProps_Init(&p->lzma2Props); +} + + +static void XzEncProps_Normalize_Fixed(CXzProps *p) +{ + UInt64 fileSize; + int t1, t1n, t2, t2r, t3; + { + CLzma2EncProps tp = p->lzma2Props; + if (tp.numTotalThreads <= 0) + tp.numTotalThreads = p->numTotalThreads; + Lzma2EncProps_Normalize(&tp); + t1n = tp.numTotalThreads; + } + + t1 = p->lzma2Props.numTotalThreads; + t2 = p->numBlockThreads_Max; + t3 = p->numTotalThreads; + + if (t2 > MTCODER__THREADS_MAX) + t2 = MTCODER__THREADS_MAX; + + if (t3 <= 0) + { + if (t2 <= 0) + t2 = 1; + t3 = t1n * t2; + } + else if (t2 <= 0) + { + t2 = t3 / t1n; + if (t2 == 0) + { + t1 = 1; + t2 = t3; + } + if (t2 > MTCODER__THREADS_MAX) + t2 = MTCODER__THREADS_MAX; + } + else if (t1 <= 0) + { + t1 = t3 / t2; + if (t1 == 0) + t1 = 1; + } + else + t3 = t1n * t2; + + p->lzma2Props.numTotalThreads = t1; + + t2r = t2; + + fileSize = p->reduceSize; + + if ((p->blockSize < fileSize || fileSize == (UInt64)(Int64)-1)) + p->lzma2Props.lzmaProps.reduceSize = p->blockSize; + + Lzma2EncProps_Normalize(&p->lzma2Props); + + t1 = p->lzma2Props.numTotalThreads; + + { + if (t2 > 1 && fileSize != (UInt64)(Int64)-1) + { + UInt64 numBlocks = fileSize / p->blockSize; + if (numBlocks * p->blockSize != fileSize) + numBlocks++; + if (numBlocks < (unsigned)t2) + { + t2r = (int)numBlocks; + if (t2r == 0) + t2r = 1; + t3 = t1 * t2r; + } + } + } + + p->numBlockThreads_Max = t2; + p->numBlockThreads_Reduced = t2r; + p->numTotalThreads = t3; +} + + +static void XzProps_Normalize(CXzProps *p) +{ + /* we normalize xzProps properties, but we normalize only some of CXzProps::lzma2Props properties. + Lzma2Enc_SetProps() will normalize lzma2Props later. */ + + if (p->blockSize == XZ_PROPS__BLOCK_SIZE__SOLID) + { + p->lzma2Props.lzmaProps.reduceSize = p->reduceSize; + p->numBlockThreads_Reduced = 1; + p->numBlockThreads_Max = 1; + if (p->lzma2Props.numTotalThreads <= 0) + p->lzma2Props.numTotalThreads = p->numTotalThreads; + return; + } + else + { + CLzma2EncProps *lzma2 = &p->lzma2Props; + if (p->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO) + { + // xz-auto + p->lzma2Props.lzmaProps.reduceSize = p->reduceSize; + + if (lzma2->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID) + { + // if (xz-auto && lzma2-solid) - we use solid for both + p->blockSize = XZ_PROPS__BLOCK_SIZE__SOLID; + p->numBlockThreads_Reduced = 1; + p->numBlockThreads_Max = 1; + if (p->lzma2Props.numTotalThreads <= 0) + p->lzma2Props.numTotalThreads = p->numTotalThreads; + } + else + { + // if (xz-auto && (lzma2-auto || lzma2-fixed_) + // we calculate block size for lzma2 and use that block size for xz, lzma2 uses single-chunk per block + CLzma2EncProps tp = p->lzma2Props; + if (tp.numTotalThreads <= 0) + tp.numTotalThreads = p->numTotalThreads; + + Lzma2EncProps_Normalize(&tp); + + p->blockSize = tp.blockSize; // fixed or solid + p->numBlockThreads_Reduced = tp.numBlockThreads_Reduced; + p->numBlockThreads_Max = tp.numBlockThreads_Max; + if (lzma2->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO) + lzma2->blockSize = tp.blockSize; // fixed or solid, LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID + if (lzma2->lzmaProps.reduceSize > tp.blockSize && tp.blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID) + lzma2->lzmaProps.reduceSize = tp.blockSize; + lzma2->numBlockThreads_Reduced = 1; + lzma2->numBlockThreads_Max = 1; + return; + } + } + else + { + // xz-fixed + // we can use xz::reduceSize or xz::blockSize as base for lzmaProps::reduceSize + + p->lzma2Props.lzmaProps.reduceSize = p->reduceSize; + { + UInt64 r = p->reduceSize; + if (r > p->blockSize || r == (UInt64)(Int64)-1) + r = p->blockSize; + lzma2->lzmaProps.reduceSize = r; + } + if (lzma2->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO) + lzma2->blockSize = LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID; + else if (lzma2->blockSize > p->blockSize && lzma2->blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID) + lzma2->blockSize = p->blockSize; + + XzEncProps_Normalize_Fixed(p); + } + } +} + + +/* ---------- CLzma2WithFilters ---------- */ + +typedef struct +{ + CLzma2EncHandle lzma2; + CSeqInFilter filter; + + #ifdef USE_SUBBLOCK + CSbEncInStream sb; + #endif +} CLzma2WithFilters; + + +static void Lzma2WithFilters_Construct(CLzma2WithFilters *p) +{ + p->lzma2 = NULL; + SeqInFilter_Construct(&p->filter); + + #ifdef USE_SUBBLOCK + SbEncInStream_Construct(&p->sb, alloc); + #endif +} + + +static SRes Lzma2WithFilters_Create(CLzma2WithFilters *p, ISzAllocPtr alloc, ISzAllocPtr bigAlloc) +{ + if (!p->lzma2) + { + p->lzma2 = Lzma2Enc_Create(alloc, bigAlloc); + if (!p->lzma2) + return SZ_ERROR_MEM; + } + return SZ_OK; +} + + +static void Lzma2WithFilters_Free(CLzma2WithFilters *p, ISzAllocPtr alloc) +{ + #ifdef USE_SUBBLOCK + SbEncInStream_Free(&p->sb); + #endif + + SeqInFilter_Free(&p->filter, alloc); + if (p->lzma2) + { + Lzma2Enc_Destroy(p->lzma2); + p->lzma2 = NULL; + } +} + + +typedef struct +{ + UInt64 unpackSize; + UInt64 totalSize; + size_t headerSize; +} CXzEncBlockInfo; + + +static SRes Xz_CompressBlock( + CLzma2WithFilters *lzmaf, + + ISeqOutStream *outStream, + Byte *outBufHeader, + Byte *outBufData, size_t outBufDataLimit, + + ISeqInStream *inStream, + // UInt64 expectedSize, + const Byte *inBuf, // used if (!inStream) + size_t inBufSize, // used if (!inStream), it's block size, props->blockSize is ignored + + const CXzProps *props, + ICompressProgress *progress, + int *inStreamFinished, /* only for inStream version */ + CXzEncBlockInfo *blockSizes, + ISzAllocPtr alloc, + ISzAllocPtr allocBig) +{ + CSeqCheckInStream checkInStream; + CSeqSizeOutStream seqSizeOutStream; + CXzBlock block; + unsigned filterIndex = 0; + CXzFilter *filter = NULL; + const CXzFilterProps *fp = &props->filterProps; + if (fp->id == 0) + fp = NULL; + + *inStreamFinished = False; + + RINOK(Lzma2WithFilters_Create(lzmaf, alloc, allocBig)); + + RINOK(Lzma2Enc_SetProps(lzmaf->lzma2, &props->lzma2Props)); + + XzBlock_ClearFlags(&block); + XzBlock_SetNumFilters(&block, 1 + (fp ? 1 : 0)); + + if (fp) + { + filter = &block.filters[filterIndex++]; + filter->id = fp->id; + filter->propsSize = 0; + + if (fp->id == XZ_ID_Delta) + { + filter->props[0] = (Byte)(fp->delta - 1); + filter->propsSize = 1; + } + else if (fp->ipDefined) + { + Byte *ptr = filter->props; + SetUi32(ptr, fp->ip); + filter->propsSize = 4; + } + } + + { + CXzFilter *f = &block.filters[filterIndex++]; + f->id = XZ_ID_LZMA2; + f->propsSize = 1; + f->props[0] = Lzma2Enc_WriteProperties(lzmaf->lzma2); + } + + seqSizeOutStream.vt.Write = SeqSizeOutStream_Write; + seqSizeOutStream.realStream = outStream; + seqSizeOutStream.outBuf = outBufData; + seqSizeOutStream.outBufLimit = outBufDataLimit; + seqSizeOutStream.processed = 0; + + /* + if (expectedSize != (UInt64)(Int64)-1) + { + block.unpackSize = expectedSize; + if (props->blockSize != (UInt64)(Int64)-1) + if (expectedSize > props->blockSize) + block.unpackSize = props->blockSize; + XzBlock_SetHasUnpackSize(&block); + } + */ + + if (outStream) + { + RINOK(XzBlock_WriteHeader(&block, &seqSizeOutStream.vt)); + } + + checkInStream.vt.Read = SeqCheckInStream_Read; + SeqCheckInStream_Init(&checkInStream, props->checkId); + + checkInStream.realStream = inStream; + checkInStream.data = inBuf; + checkInStream.limit = props->blockSize; + if (!inStream) + checkInStream.limit = inBufSize; + + if (fp) + { + #ifdef USE_SUBBLOCK + if (fp->id == XZ_ID_Subblock) + { + lzmaf->sb.inStream = &checkInStream.vt; + RINOK(SbEncInStream_Init(&lzmaf->sb)); + } + else + #endif + { + lzmaf->filter.realStream = &checkInStream.vt; + RINOK(SeqInFilter_Init(&lzmaf->filter, filter, alloc)); + } + } + + { + SRes res; + Byte *outBuf = NULL; + size_t outSize = 0; + BoolInt useStream = (fp || inStream); + // useStream = True; + + if (!useStream) + { + XzCheck_Update(&checkInStream.check, inBuf, inBufSize); + checkInStream.processed = inBufSize; + } + + if (!outStream) + { + outBuf = seqSizeOutStream.outBuf; // + (size_t)seqSizeOutStream.processed; + outSize = seqSizeOutStream.outBufLimit; // - (size_t)seqSizeOutStream.processed; + } + + res = Lzma2Enc_Encode2(lzmaf->lzma2, + outBuf ? NULL : &seqSizeOutStream.vt, + outBuf, + outBuf ? &outSize : NULL, + + useStream ? + (fp ? + ( + #ifdef USE_SUBBLOCK + (fp->id == XZ_ID_Subblock) ? &lzmaf->sb.vt: + #endif + &lzmaf->filter.p) : + &checkInStream.vt) : NULL, + + useStream ? NULL : inBuf, + useStream ? 0 : inBufSize, + + progress); + + if (outBuf) + seqSizeOutStream.processed += outSize; + + RINOK(res); + blockSizes->unpackSize = checkInStream.processed; + } + { + Byte buf[4 + 64]; + unsigned padSize = XZ_GET_PAD_SIZE(seqSizeOutStream.processed); + UInt64 packSize = seqSizeOutStream.processed; + + buf[0] = 0; + buf[1] = 0; + buf[2] = 0; + buf[3] = 0; + + SeqCheckInStream_GetDigest(&checkInStream, buf + 4); + RINOK(WriteBytes(&seqSizeOutStream.vt, buf + (4 - padSize), padSize + XzFlags_GetCheckSize((CXzStreamFlags)props->checkId))); + + blockSizes->totalSize = seqSizeOutStream.processed - padSize; + + if (!outStream) + { + seqSizeOutStream.outBuf = outBufHeader; + seqSizeOutStream.outBufLimit = XZ_BLOCK_HEADER_SIZE_MAX; + seqSizeOutStream.processed = 0; + + block.unpackSize = blockSizes->unpackSize; + XzBlock_SetHasUnpackSize(&block); + + block.packSize = packSize; + XzBlock_SetHasPackSize(&block); + + RINOK(XzBlock_WriteHeader(&block, &seqSizeOutStream.vt)); + + blockSizes->headerSize = (size_t)seqSizeOutStream.processed; + blockSizes->totalSize += seqSizeOutStream.processed; + } + } + + if (inStream) + *inStreamFinished = checkInStream.realStreamFinished; + else + { + *inStreamFinished = False; + if (checkInStream.processed != inBufSize) + return SZ_ERROR_FAIL; + } + + return SZ_OK; +} + + + +typedef struct +{ + ICompressProgress vt; + ICompressProgress *progress; + UInt64 inOffset; + UInt64 outOffset; +} CCompressProgress_XzEncOffset; + + +static SRes CompressProgress_XzEncOffset_Progress(const ICompressProgress *pp, UInt64 inSize, UInt64 outSize) +{ + const CCompressProgress_XzEncOffset *p = CONTAINER_FROM_VTBL(pp, CCompressProgress_XzEncOffset, vt); + inSize += p->inOffset; + outSize += p->outOffset; + return ICompressProgress_Progress(p->progress, inSize, outSize); +} + + + + +typedef struct +{ + ISzAllocPtr alloc; + ISzAllocPtr allocBig; + + CXzProps xzProps; + UInt64 expectedDataSize; + + CXzEncIndex xzIndex; + + CLzma2WithFilters lzmaf_Items[MTCODER__THREADS_MAX]; + + size_t outBufSize; /* size of allocated outBufs[i] */ + Byte *outBufs[MTCODER__BLOCKS_MAX]; + + #ifndef _7ZIP_ST + unsigned checkType; + ISeqOutStream *outStream; + BoolInt mtCoder_WasConstructed; + CMtCoder mtCoder; + CXzEncBlockInfo EncBlocks[MTCODER__BLOCKS_MAX]; + #endif + +} CXzEnc; + + +static void XzEnc_Construct(CXzEnc *p) +{ + unsigned i; + + XzEncIndex_Construct(&p->xzIndex); + + for (i = 0; i < MTCODER__THREADS_MAX; i++) + Lzma2WithFilters_Construct(&p->lzmaf_Items[i]); + + #ifndef _7ZIP_ST + p->mtCoder_WasConstructed = False; + { + for (i = 0; i < MTCODER__BLOCKS_MAX; i++) + p->outBufs[i] = NULL; + p->outBufSize = 0; + } + #endif +} + + +static void XzEnc_FreeOutBufs(CXzEnc *p) +{ + unsigned i; + for (i = 0; i < MTCODER__BLOCKS_MAX; i++) + if (p->outBufs[i]) + { + ISzAlloc_Free(p->alloc, p->outBufs[i]); + p->outBufs[i] = NULL; + } + p->outBufSize = 0; +} + + +static void XzEnc_Free(CXzEnc *p, ISzAllocPtr alloc) +{ + unsigned i; + + XzEncIndex_Free(&p->xzIndex, alloc); + + for (i = 0; i < MTCODER__THREADS_MAX; i++) + Lzma2WithFilters_Free(&p->lzmaf_Items[i], alloc); + + #ifndef _7ZIP_ST + if (p->mtCoder_WasConstructed) + { + MtCoder_Destruct(&p->mtCoder); + p->mtCoder_WasConstructed = False; + } + XzEnc_FreeOutBufs(p); + #endif +} + + +CXzEncHandle XzEnc_Create(ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CXzEnc *p = (CXzEnc *)ISzAlloc_Alloc(alloc, sizeof(CXzEnc)); + if (!p) + return NULL; + XzEnc_Construct(p); + XzProps_Init(&p->xzProps); + XzProps_Normalize(&p->xzProps); + p->expectedDataSize = (UInt64)(Int64)-1; + p->alloc = alloc; + p->allocBig = allocBig; + return p; +} + + +void XzEnc_Destroy(CXzEncHandle pp) +{ + CXzEnc *p = (CXzEnc *)pp; + XzEnc_Free(p, p->alloc); + ISzAlloc_Free(p->alloc, p); +} + + +SRes XzEnc_SetProps(CXzEncHandle pp, const CXzProps *props) +{ + CXzEnc *p = (CXzEnc *)pp; + p->xzProps = *props; + XzProps_Normalize(&p->xzProps); + return SZ_OK; +} + + +void XzEnc_SetDataSize(CXzEncHandle pp, UInt64 expectedDataSiize) +{ + CXzEnc *p = (CXzEnc *)pp; + p->expectedDataSize = expectedDataSiize; +} + + + + +#ifndef _7ZIP_ST + +static SRes XzEnc_MtCallback_Code(void *pp, unsigned coderIndex, unsigned outBufIndex, + const Byte *src, size_t srcSize, int finished) +{ + CXzEnc *me = (CXzEnc *)pp; + SRes res; + CMtProgressThunk progressThunk; + + Byte *dest = me->outBufs[outBufIndex]; + + UNUSED_VAR(finished) + + { + CXzEncBlockInfo *bInfo = &me->EncBlocks[outBufIndex]; + bInfo->totalSize = 0; + bInfo->unpackSize = 0; + bInfo->headerSize = 0; + } + + if (!dest) + { + dest = (Byte *)ISzAlloc_Alloc(me->alloc, me->outBufSize); + if (!dest) + return SZ_ERROR_MEM; + me->outBufs[outBufIndex] = dest; + } + + MtProgressThunk_CreateVTable(&progressThunk); + progressThunk.mtProgress = &me->mtCoder.mtProgress; + MtProgressThunk_Init(&progressThunk); + + { + CXzEncBlockInfo blockSizes; + int inStreamFinished; + + res = Xz_CompressBlock( + &me->lzmaf_Items[coderIndex], + + NULL, + dest, + dest + XZ_BLOCK_HEADER_SIZE_MAX, me->outBufSize - XZ_BLOCK_HEADER_SIZE_MAX, + + NULL, + // srcSize, // expectedSize + src, srcSize, + + &me->xzProps, + &progressThunk.vt, + &inStreamFinished, + &blockSizes, + me->alloc, + me->allocBig); + + if (res == SZ_OK) + me->EncBlocks[outBufIndex] = blockSizes; + + return res; + } +} + + +static SRes XzEnc_MtCallback_Write(void *pp, unsigned outBufIndex) +{ + CXzEnc *me = (CXzEnc *)pp; + + const CXzEncBlockInfo *bInfo = &me->EncBlocks[outBufIndex]; + const Byte *data = me->outBufs[outBufIndex]; + + RINOK(WriteBytes(me->outStream, data, bInfo->headerSize)); + + { + UInt64 totalPackFull = bInfo->totalSize + XZ_GET_PAD_SIZE(bInfo->totalSize); + RINOK(WriteBytes(me->outStream, data + XZ_BLOCK_HEADER_SIZE_MAX, (size_t)totalPackFull - bInfo->headerSize)); + } + + return XzEncIndex_AddIndexRecord(&me->xzIndex, bInfo->unpackSize, bInfo->totalSize, me->alloc); +} + +#endif + + + +SRes XzEnc_Encode(CXzEncHandle pp, ISeqOutStream *outStream, ISeqInStream *inStream, ICompressProgress *progress) +{ + CXzEnc *p = (CXzEnc *)pp; + + const CXzProps *props = &p->xzProps; + + XzEncIndex_Init(&p->xzIndex); + { + UInt64 numBlocks = 1; + UInt64 blockSize = props->blockSize; + + if (blockSize != XZ_PROPS__BLOCK_SIZE__SOLID + && props->reduceSize != (UInt64)(Int64)-1) + { + numBlocks = props->reduceSize / blockSize; + if (numBlocks * blockSize != props->reduceSize) + numBlocks++; + } + else + blockSize = (UInt64)1 << 62; + + RINOK(XzEncIndex_PreAlloc(&p->xzIndex, numBlocks, blockSize, XZ_GET_ESTIMATED_BLOCK_TOTAL_PACK_SIZE(blockSize), p->alloc)); + } + + RINOK(Xz_WriteHeader((CXzStreamFlags)props->checkId, outStream)); + + + #ifndef _7ZIP_ST + if (props->numBlockThreads_Reduced > 1) + { + IMtCoderCallback2 vt; + + if (!p->mtCoder_WasConstructed) + { + p->mtCoder_WasConstructed = True; + MtCoder_Construct(&p->mtCoder); + } + + vt.Code = XzEnc_MtCallback_Code; + vt.Write = XzEnc_MtCallback_Write; + + p->checkType = props->checkId; + p->xzProps = *props; + + p->outStream = outStream; + + p->mtCoder.allocBig = p->allocBig; + p->mtCoder.progress = progress; + p->mtCoder.inStream = inStream; + p->mtCoder.inData = NULL; + p->mtCoder.inDataSize = 0; + p->mtCoder.mtCallback = &vt; + p->mtCoder.mtCallbackObject = p; + + if ( props->blockSize == XZ_PROPS__BLOCK_SIZE__SOLID + || props->blockSize == XZ_PROPS__BLOCK_SIZE__AUTO) + return SZ_ERROR_FAIL; + + p->mtCoder.blockSize = (size_t)props->blockSize; + if (p->mtCoder.blockSize != props->blockSize) + return SZ_ERROR_PARAM; /* SZ_ERROR_MEM */ + + { + size_t destBlockSize = XZ_BLOCK_HEADER_SIZE_MAX + XZ_GET_MAX_BLOCK_PACK_SIZE(p->mtCoder.blockSize); + if (destBlockSize < p->mtCoder.blockSize) + return SZ_ERROR_PARAM; + if (p->outBufSize != destBlockSize) + XzEnc_FreeOutBufs(p); + p->outBufSize = destBlockSize; + } + + p->mtCoder.numThreadsMax = (unsigned)props->numBlockThreads_Max; + p->mtCoder.expectedDataSize = p->expectedDataSize; + + RINOK(MtCoder_Code(&p->mtCoder)); + } + else + #endif + { + int writeStartSizes; + CCompressProgress_XzEncOffset progress2; + Byte *bufData = NULL; + size_t bufSize = 0; + + progress2.vt.Progress = CompressProgress_XzEncOffset_Progress; + progress2.inOffset = 0; + progress2.outOffset = 0; + progress2.progress = progress; + + writeStartSizes = 0; + + if (props->blockSize != XZ_PROPS__BLOCK_SIZE__SOLID) + { + writeStartSizes = (props->forceWriteSizesInHeader > 0); + + if (writeStartSizes) + { + size_t t2; + size_t t = (size_t)props->blockSize; + if (t != props->blockSize) + return SZ_ERROR_PARAM; + t = XZ_GET_MAX_BLOCK_PACK_SIZE(t); + if (t < props->blockSize) + return SZ_ERROR_PARAM; + t2 = XZ_BLOCK_HEADER_SIZE_MAX + t; + if (!p->outBufs[0] || t2 != p->outBufSize) + { + XzEnc_FreeOutBufs(p); + p->outBufs[0] = (Byte *)ISzAlloc_Alloc(p->alloc, t2); + if (!p->outBufs[0]) + return SZ_ERROR_MEM; + p->outBufSize = t2; + } + bufData = p->outBufs[0] + XZ_BLOCK_HEADER_SIZE_MAX; + bufSize = t; + } + } + + for (;;) + { + CXzEncBlockInfo blockSizes; + int inStreamFinished; + + /* + UInt64 rem = (UInt64)(Int64)-1; + if (props->reduceSize != (UInt64)(Int64)-1 + && props->reduceSize >= progress2.inOffset) + rem = props->reduceSize - progress2.inOffset; + */ + + blockSizes.headerSize = 0; // for GCC + + RINOK(Xz_CompressBlock( + &p->lzmaf_Items[0], + + writeStartSizes ? NULL : outStream, + writeStartSizes ? p->outBufs[0] : NULL, + bufData, bufSize, + + inStream, + // rem, + NULL, 0, + + props, + progress ? &progress2.vt : NULL, + &inStreamFinished, + &blockSizes, + p->alloc, + p->allocBig)); + + { + UInt64 totalPackFull = blockSizes.totalSize + XZ_GET_PAD_SIZE(blockSizes.totalSize); + + if (writeStartSizes) + { + RINOK(WriteBytes(outStream, p->outBufs[0], blockSizes.headerSize)); + RINOK(WriteBytes(outStream, bufData, (size_t)totalPackFull - blockSizes.headerSize)); + } + + RINOK(XzEncIndex_AddIndexRecord(&p->xzIndex, blockSizes.unpackSize, blockSizes.totalSize, p->alloc)); + + progress2.inOffset += blockSizes.unpackSize; + progress2.outOffset += totalPackFull; + } + + if (inStreamFinished) + break; + } + } + + return XzEncIndex_WriteFooter(&p->xzIndex, (CXzStreamFlags)props->checkId, outStream); +} + + +#include "Alloc.h" + +SRes Xz_Encode(ISeqOutStream *outStream, ISeqInStream *inStream, + const CXzProps *props, ICompressProgress *progress) +{ + SRes res; + CXzEncHandle xz = XzEnc_Create(&g_Alloc, &g_BigAlloc); + if (!xz) + return SZ_ERROR_MEM; + res = XzEnc_SetProps(xz, props); + if (res == SZ_OK) + res = XzEnc_Encode(xz, outStream, inStream, progress); + XzEnc_Destroy(xz); + return res; +} + + +SRes Xz_EncodeEmpty(ISeqOutStream *outStream) +{ + SRes res; + CXzEncIndex xzIndex; + XzEncIndex_Construct(&xzIndex); + res = Xz_WriteHeader((CXzStreamFlags)0, outStream); + if (res == SZ_OK) + res = XzEncIndex_WriteFooter(&xzIndex, (CXzStreamFlags)0, outStream); + XzEncIndex_Free(&xzIndex, NULL); // g_Alloc + return res; +} diff --git a/deps/LZMA-SDK/C/XzEnc.h b/deps/LZMA-SDK/C/XzEnc.h index 529ac3fd8..0c29e7e1e 100644 --- a/deps/LZMA-SDK/C/XzEnc.h +++ b/deps/LZMA-SDK/C/XzEnc.h @@ -1,60 +1,60 @@ -/* XzEnc.h -- Xz Encode -2017-06-27 : Igor Pavlov : Public domain */ - -#ifndef __XZ_ENC_H -#define __XZ_ENC_H - -#include "Lzma2Enc.h" - -#include "Xz.h" - -EXTERN_C_BEGIN - - -#define XZ_PROPS__BLOCK_SIZE__AUTO LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO -#define XZ_PROPS__BLOCK_SIZE__SOLID LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID - - -typedef struct -{ - UInt32 id; - UInt32 delta; - UInt32 ip; - int ipDefined; -} CXzFilterProps; - -void XzFilterProps_Init(CXzFilterProps *p); - - -typedef struct -{ - CLzma2EncProps lzma2Props; - CXzFilterProps filterProps; - unsigned checkId; - UInt64 blockSize; - int numBlockThreads_Reduced; - int numBlockThreads_Max; - int numTotalThreads; - int forceWriteSizesInHeader; - UInt64 reduceSize; -} CXzProps; - -void XzProps_Init(CXzProps *p); - - -typedef void * CXzEncHandle; - -CXzEncHandle XzEnc_Create(ISzAllocPtr alloc, ISzAllocPtr allocBig); -void XzEnc_Destroy(CXzEncHandle p); -SRes XzEnc_SetProps(CXzEncHandle p, const CXzProps *props); -void XzEnc_SetDataSize(CXzEncHandle p, UInt64 expectedDataSiize); -SRes XzEnc_Encode(CXzEncHandle p, ISeqOutStream *outStream, ISeqInStream *inStream, ICompressProgress *progress); - -SRes Xz_Encode(ISeqOutStream *outStream, ISeqInStream *inStream, - const CXzProps *props, ICompressProgress *progress); - -SRes Xz_EncodeEmpty(ISeqOutStream *outStream); - -EXTERN_C_END - -#endif +/* XzEnc.h -- Xz Encode +2017-06-27 : Igor Pavlov : Public domain */ + +#ifndef __XZ_ENC_H +#define __XZ_ENC_H + +#include "Lzma2Enc.h" + +#include "Xz.h" + +EXTERN_C_BEGIN + + +#define XZ_PROPS__BLOCK_SIZE__AUTO LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO +#define XZ_PROPS__BLOCK_SIZE__SOLID LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID + + +typedef struct +{ + UInt32 id; + UInt32 delta; + UInt32 ip; + int ipDefined; +} CXzFilterProps; + +void XzFilterProps_Init(CXzFilterProps *p); + + +typedef struct +{ + CLzma2EncProps lzma2Props; + CXzFilterProps filterProps; + unsigned checkId; + UInt64 blockSize; + int numBlockThreads_Reduced; + int numBlockThreads_Max; + int numTotalThreads; + int forceWriteSizesInHeader; + UInt64 reduceSize; +} CXzProps; + +void XzProps_Init(CXzProps *p); + + +typedef void * CXzEncHandle; + +CXzEncHandle XzEnc_Create(ISzAllocPtr alloc, ISzAllocPtr allocBig); +void XzEnc_Destroy(CXzEncHandle p); +SRes XzEnc_SetProps(CXzEncHandle p, const CXzProps *props); +void XzEnc_SetDataSize(CXzEncHandle p, UInt64 expectedDataSiize); +SRes XzEnc_Encode(CXzEncHandle p, ISeqOutStream *outStream, ISeqInStream *inStream, ICompressProgress *progress); + +SRes Xz_Encode(ISeqOutStream *outStream, ISeqInStream *inStream, + const CXzProps *props, ICompressProgress *progress); + +SRes Xz_EncodeEmpty(ISeqOutStream *outStream); + +EXTERN_C_END + +#endif diff --git a/deps/LZMA-SDK/C/XzIn.c b/deps/LZMA-SDK/C/XzIn.c index 07201b842..84f868ec6 100644 --- a/deps/LZMA-SDK/C/XzIn.c +++ b/deps/LZMA-SDK/C/XzIn.c @@ -1,325 +1,325 @@ -/* XzIn.c - Xz input -2021-09-04 : Igor Pavlov : Public domain */ - -#include "Precomp.h" - -#include - -#include "7zCrc.h" -#include "CpuArch.h" -#include "Xz.h" - -/* -#define XZ_FOOTER_SIG_CHECK(p) (memcmp((p), XZ_FOOTER_SIG, XZ_FOOTER_SIG_SIZE) == 0) -*/ -#define XZ_FOOTER_SIG_CHECK(p) ((p)[0] == XZ_FOOTER_SIG_0 && (p)[1] == XZ_FOOTER_SIG_1) - - -SRes Xz_ReadHeader(CXzStreamFlags *p, ISeqInStream *inStream) -{ - Byte sig[XZ_STREAM_HEADER_SIZE]; - RINOK(SeqInStream_Read2(inStream, sig, XZ_STREAM_HEADER_SIZE, SZ_ERROR_NO_ARCHIVE)); - if (memcmp(sig, XZ_SIG, XZ_SIG_SIZE) != 0) - return SZ_ERROR_NO_ARCHIVE; - return Xz_ParseHeader(p, sig); -} - -#define READ_VARINT_AND_CHECK(buf, pos, size, res) \ - { unsigned s = Xz_ReadVarInt(buf + pos, size - pos, res); \ - if (s == 0) return SZ_ERROR_ARCHIVE; \ - pos += s; } - -SRes XzBlock_ReadHeader(CXzBlock *p, ISeqInStream *inStream, BoolInt *isIndex, UInt32 *headerSizeRes) -{ - Byte header[XZ_BLOCK_HEADER_SIZE_MAX]; - unsigned headerSize; - *headerSizeRes = 0; - RINOK(SeqInStream_ReadByte(inStream, &header[0])); - headerSize = (unsigned)header[0]; - if (headerSize == 0) - { - *headerSizeRes = 1; - *isIndex = True; - return SZ_OK; - } - - *isIndex = False; - headerSize = (headerSize << 2) + 4; - *headerSizeRes = headerSize; - RINOK(SeqInStream_Read(inStream, header + 1, headerSize - 1)); - return XzBlock_Parse(p, header); -} - -#define ADD_SIZE_CHECK(size, val) \ - { UInt64 newSize = size + (val); if (newSize < size) return XZ_SIZE_OVERFLOW; size = newSize; } - -UInt64 Xz_GetUnpackSize(const CXzStream *p) -{ - UInt64 size = 0; - size_t i; - for (i = 0; i < p->numBlocks; i++) - ADD_SIZE_CHECK(size, p->blocks[i].unpackSize); - return size; -} - -UInt64 Xz_GetPackSize(const CXzStream *p) -{ - UInt64 size = 0; - size_t i; - for (i = 0; i < p->numBlocks; i++) - ADD_SIZE_CHECK(size, (p->blocks[i].totalSize + 3) & ~(UInt64)3); - return size; -} - -/* -SRes XzBlock_ReadFooter(CXzBlock *p, CXzStreamFlags f, ISeqInStream *inStream) -{ - return SeqInStream_Read(inStream, p->check, XzFlags_GetCheckSize(f)); -} -*/ - -static SRes Xz_ReadIndex2(CXzStream *p, const Byte *buf, size_t size, ISzAllocPtr alloc) -{ - size_t numBlocks, pos = 1; - UInt32 crc; - - if (size < 5 || buf[0] != 0) - return SZ_ERROR_ARCHIVE; - - size -= 4; - crc = CrcCalc(buf, size); - if (crc != GetUi32(buf + size)) - return SZ_ERROR_ARCHIVE; - - { - UInt64 numBlocks64; - READ_VARINT_AND_CHECK(buf, pos, size, &numBlocks64); - numBlocks = (size_t)numBlocks64; - if (numBlocks != numBlocks64 || numBlocks * 2 > size) - return SZ_ERROR_ARCHIVE; - } - - Xz_Free(p, alloc); - if (numBlocks != 0) - { - size_t i; - p->numBlocks = numBlocks; - p->blocks = (CXzBlockSizes *)ISzAlloc_Alloc(alloc, sizeof(CXzBlockSizes) * numBlocks); - if (!p->blocks) - return SZ_ERROR_MEM; - for (i = 0; i < numBlocks; i++) - { - CXzBlockSizes *block = &p->blocks[i]; - READ_VARINT_AND_CHECK(buf, pos, size, &block->totalSize); - READ_VARINT_AND_CHECK(buf, pos, size, &block->unpackSize); - if (block->totalSize == 0) - return SZ_ERROR_ARCHIVE; - } - } - while ((pos & 3) != 0) - if (buf[pos++] != 0) - return SZ_ERROR_ARCHIVE; - return (pos == size) ? SZ_OK : SZ_ERROR_ARCHIVE; -} - -static SRes Xz_ReadIndex(CXzStream *p, ILookInStream *stream, UInt64 indexSize, ISzAllocPtr alloc) -{ - SRes res; - size_t size; - Byte *buf; - if (indexSize > ((UInt32)1 << 31)) - return SZ_ERROR_UNSUPPORTED; - size = (size_t)indexSize; - if (size != indexSize) - return SZ_ERROR_UNSUPPORTED; - buf = (Byte *)ISzAlloc_Alloc(alloc, size); - if (!buf) - return SZ_ERROR_MEM; - res = LookInStream_Read2(stream, buf, size, SZ_ERROR_UNSUPPORTED); - if (res == SZ_OK) - res = Xz_ReadIndex2(p, buf, size, alloc); - ISzAlloc_Free(alloc, buf); - return res; -} - -static SRes LookInStream_SeekRead_ForArc(ILookInStream *stream, UInt64 offset, void *buf, size_t size) -{ - RINOK(LookInStream_SeekTo(stream, offset)); - return LookInStream_Read(stream, buf, size); - /* return LookInStream_Read2(stream, buf, size, SZ_ERROR_NO_ARCHIVE); */ -} - -static SRes Xz_ReadBackward(CXzStream *p, ILookInStream *stream, Int64 *startOffset, ISzAllocPtr alloc) -{ - UInt64 indexSize; - Byte buf[XZ_STREAM_FOOTER_SIZE]; - UInt64 pos = (UInt64)*startOffset; - - if ((pos & 3) != 0 || pos < XZ_STREAM_FOOTER_SIZE) - return SZ_ERROR_NO_ARCHIVE; - - pos -= XZ_STREAM_FOOTER_SIZE; - RINOK(LookInStream_SeekRead_ForArc(stream, pos, buf, XZ_STREAM_FOOTER_SIZE)); - - if (!XZ_FOOTER_SIG_CHECK(buf + 10)) - { - UInt32 total = 0; - pos += XZ_STREAM_FOOTER_SIZE; - - for (;;) - { - size_t i; - #define TEMP_BUF_SIZE (1 << 10) - Byte temp[TEMP_BUF_SIZE]; - - i = (pos > TEMP_BUF_SIZE) ? TEMP_BUF_SIZE : (size_t)pos; - pos -= i; - RINOK(LookInStream_SeekRead_ForArc(stream, pos, temp, i)); - total += (UInt32)i; - for (; i != 0; i--) - if (temp[i - 1] != 0) - break; - if (i != 0) - { - if ((i & 3) != 0) - return SZ_ERROR_NO_ARCHIVE; - pos += i; - break; - } - if (pos < XZ_STREAM_FOOTER_SIZE || total > (1 << 16)) - return SZ_ERROR_NO_ARCHIVE; - } - - if (pos < XZ_STREAM_FOOTER_SIZE) - return SZ_ERROR_NO_ARCHIVE; - pos -= XZ_STREAM_FOOTER_SIZE; - RINOK(LookInStream_SeekRead_ForArc(stream, pos, buf, XZ_STREAM_FOOTER_SIZE)); - if (!XZ_FOOTER_SIG_CHECK(buf + 10)) - return SZ_ERROR_NO_ARCHIVE; - } - - p->flags = (CXzStreamFlags)GetBe16(buf + 8); - - if (!XzFlags_IsSupported(p->flags)) - return SZ_ERROR_UNSUPPORTED; - - { - /* to eliminate GCC 6.3 warning: - dereferencing type-punned pointer will break strict-aliasing rules */ - const Byte *buf_ptr = buf; - if (GetUi32(buf_ptr) != CrcCalc(buf + 4, 6)) - return SZ_ERROR_ARCHIVE; - } - - indexSize = ((UInt64)GetUi32(buf + 4) + 1) << 2; - - if (pos < indexSize) - return SZ_ERROR_ARCHIVE; - - pos -= indexSize; - RINOK(LookInStream_SeekTo(stream, pos)); - RINOK(Xz_ReadIndex(p, stream, indexSize, alloc)); - - { - UInt64 totalSize = Xz_GetPackSize(p); - if (totalSize == XZ_SIZE_OVERFLOW - || totalSize >= ((UInt64)1 << 63) - || pos < totalSize + XZ_STREAM_HEADER_SIZE) - return SZ_ERROR_ARCHIVE; - pos -= (totalSize + XZ_STREAM_HEADER_SIZE); - RINOK(LookInStream_SeekTo(stream, pos)); - *startOffset = (Int64)pos; - } - { - CXzStreamFlags headerFlags; - CSecToRead secToRead; - SecToRead_CreateVTable(&secToRead); - secToRead.realStream = stream; - - RINOK(Xz_ReadHeader(&headerFlags, &secToRead.vt)); - return (p->flags == headerFlags) ? SZ_OK : SZ_ERROR_ARCHIVE; - } -} - - -/* ---------- Xz Streams ---------- */ - -void Xzs_Construct(CXzs *p) -{ - p->num = p->numAllocated = 0; - p->streams = 0; -} - -void Xzs_Free(CXzs *p, ISzAllocPtr alloc) -{ - size_t i; - for (i = 0; i < p->num; i++) - Xz_Free(&p->streams[i], alloc); - ISzAlloc_Free(alloc, p->streams); - p->num = p->numAllocated = 0; - p->streams = 0; -} - -UInt64 Xzs_GetNumBlocks(const CXzs *p) -{ - UInt64 num = 0; - size_t i; - for (i = 0; i < p->num; i++) - num += p->streams[i].numBlocks; - return num; -} - -UInt64 Xzs_GetUnpackSize(const CXzs *p) -{ - UInt64 size = 0; - size_t i; - for (i = 0; i < p->num; i++) - ADD_SIZE_CHECK(size, Xz_GetUnpackSize(&p->streams[i])); - return size; -} - -/* -UInt64 Xzs_GetPackSize(const CXzs *p) -{ - UInt64 size = 0; - size_t i; - for (i = 0; i < p->num; i++) - ADD_SIZE_CHECK(size, Xz_GetTotalSize(&p->streams[i])); - return size; -} -*/ - -SRes Xzs_ReadBackward(CXzs *p, ILookInStream *stream, Int64 *startOffset, ICompressProgress *progress, ISzAllocPtr alloc) -{ - Int64 endOffset = 0; - RINOK(ILookInStream_Seek(stream, &endOffset, SZ_SEEK_END)); - *startOffset = endOffset; - for (;;) - { - CXzStream st; - SRes res; - Xz_Construct(&st); - res = Xz_ReadBackward(&st, stream, startOffset, alloc); - st.startOffset = (UInt64)*startOffset; - RINOK(res); - if (p->num == p->numAllocated) - { - const size_t newNum = p->num + p->num / 4 + 1; - void *data = ISzAlloc_Alloc(alloc, newNum * sizeof(CXzStream)); - if (!data) - return SZ_ERROR_MEM; - p->numAllocated = newNum; - if (p->num != 0) - memcpy(data, p->streams, p->num * sizeof(CXzStream)); - ISzAlloc_Free(alloc, p->streams); - p->streams = (CXzStream *)data; - } - p->streams[p->num++] = st; - if (*startOffset == 0) - break; - RINOK(LookInStream_SeekTo(stream, (UInt64)*startOffset)); - if (progress && ICompressProgress_Progress(progress, (UInt64)(endOffset - *startOffset), (UInt64)(Int64)-1) != SZ_OK) - return SZ_ERROR_PROGRESS; - } - return SZ_OK; -} +/* XzIn.c - Xz input +2021-09-04 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#include "7zCrc.h" +#include "CpuArch.h" +#include "Xz.h" + +/* +#define XZ_FOOTER_SIG_CHECK(p) (memcmp((p), XZ_FOOTER_SIG, XZ_FOOTER_SIG_SIZE) == 0) +*/ +#define XZ_FOOTER_SIG_CHECK(p) ((p)[0] == XZ_FOOTER_SIG_0 && (p)[1] == XZ_FOOTER_SIG_1) + + +SRes Xz_ReadHeader(CXzStreamFlags *p, ISeqInStream *inStream) +{ + Byte sig[XZ_STREAM_HEADER_SIZE]; + RINOK(SeqInStream_Read2(inStream, sig, XZ_STREAM_HEADER_SIZE, SZ_ERROR_NO_ARCHIVE)); + if (memcmp(sig, XZ_SIG, XZ_SIG_SIZE) != 0) + return SZ_ERROR_NO_ARCHIVE; + return Xz_ParseHeader(p, sig); +} + +#define READ_VARINT_AND_CHECK(buf, pos, size, res) \ + { unsigned s = Xz_ReadVarInt(buf + pos, size - pos, res); \ + if (s == 0) return SZ_ERROR_ARCHIVE; \ + pos += s; } + +SRes XzBlock_ReadHeader(CXzBlock *p, ISeqInStream *inStream, BoolInt *isIndex, UInt32 *headerSizeRes) +{ + Byte header[XZ_BLOCK_HEADER_SIZE_MAX]; + unsigned headerSize; + *headerSizeRes = 0; + RINOK(SeqInStream_ReadByte(inStream, &header[0])); + headerSize = (unsigned)header[0]; + if (headerSize == 0) + { + *headerSizeRes = 1; + *isIndex = True; + return SZ_OK; + } + + *isIndex = False; + headerSize = (headerSize << 2) + 4; + *headerSizeRes = headerSize; + RINOK(SeqInStream_Read(inStream, header + 1, headerSize - 1)); + return XzBlock_Parse(p, header); +} + +#define ADD_SIZE_CHECK(size, val) \ + { UInt64 newSize = size + (val); if (newSize < size) return XZ_SIZE_OVERFLOW; size = newSize; } + +UInt64 Xz_GetUnpackSize(const CXzStream *p) +{ + UInt64 size = 0; + size_t i; + for (i = 0; i < p->numBlocks; i++) + ADD_SIZE_CHECK(size, p->blocks[i].unpackSize); + return size; +} + +UInt64 Xz_GetPackSize(const CXzStream *p) +{ + UInt64 size = 0; + size_t i; + for (i = 0; i < p->numBlocks; i++) + ADD_SIZE_CHECK(size, (p->blocks[i].totalSize + 3) & ~(UInt64)3); + return size; +} + +/* +SRes XzBlock_ReadFooter(CXzBlock *p, CXzStreamFlags f, ISeqInStream *inStream) +{ + return SeqInStream_Read(inStream, p->check, XzFlags_GetCheckSize(f)); +} +*/ + +static SRes Xz_ReadIndex2(CXzStream *p, const Byte *buf, size_t size, ISzAllocPtr alloc) +{ + size_t numBlocks, pos = 1; + UInt32 crc; + + if (size < 5 || buf[0] != 0) + return SZ_ERROR_ARCHIVE; + + size -= 4; + crc = CrcCalc(buf, size); + if (crc != GetUi32(buf + size)) + return SZ_ERROR_ARCHIVE; + + { + UInt64 numBlocks64; + READ_VARINT_AND_CHECK(buf, pos, size, &numBlocks64); + numBlocks = (size_t)numBlocks64; + if (numBlocks != numBlocks64 || numBlocks * 2 > size) + return SZ_ERROR_ARCHIVE; + } + + Xz_Free(p, alloc); + if (numBlocks != 0) + { + size_t i; + p->numBlocks = numBlocks; + p->blocks = (CXzBlockSizes *)ISzAlloc_Alloc(alloc, sizeof(CXzBlockSizes) * numBlocks); + if (!p->blocks) + return SZ_ERROR_MEM; + for (i = 0; i < numBlocks; i++) + { + CXzBlockSizes *block = &p->blocks[i]; + READ_VARINT_AND_CHECK(buf, pos, size, &block->totalSize); + READ_VARINT_AND_CHECK(buf, pos, size, &block->unpackSize); + if (block->totalSize == 0) + return SZ_ERROR_ARCHIVE; + } + } + while ((pos & 3) != 0) + if (buf[pos++] != 0) + return SZ_ERROR_ARCHIVE; + return (pos == size) ? SZ_OK : SZ_ERROR_ARCHIVE; +} + +static SRes Xz_ReadIndex(CXzStream *p, ILookInStream *stream, UInt64 indexSize, ISzAllocPtr alloc) +{ + SRes res; + size_t size; + Byte *buf; + if (indexSize > ((UInt32)1 << 31)) + return SZ_ERROR_UNSUPPORTED; + size = (size_t)indexSize; + if (size != indexSize) + return SZ_ERROR_UNSUPPORTED; + buf = (Byte *)ISzAlloc_Alloc(alloc, size); + if (!buf) + return SZ_ERROR_MEM; + res = LookInStream_Read2(stream, buf, size, SZ_ERROR_UNSUPPORTED); + if (res == SZ_OK) + res = Xz_ReadIndex2(p, buf, size, alloc); + ISzAlloc_Free(alloc, buf); + return res; +} + +static SRes LookInStream_SeekRead_ForArc(ILookInStream *stream, UInt64 offset, void *buf, size_t size) +{ + RINOK(LookInStream_SeekTo(stream, offset)); + return LookInStream_Read(stream, buf, size); + /* return LookInStream_Read2(stream, buf, size, SZ_ERROR_NO_ARCHIVE); */ +} + +static SRes Xz_ReadBackward(CXzStream *p, ILookInStream *stream, Int64 *startOffset, ISzAllocPtr alloc) +{ + UInt64 indexSize; + Byte buf[XZ_STREAM_FOOTER_SIZE]; + UInt64 pos = (UInt64)*startOffset; + + if ((pos & 3) != 0 || pos < XZ_STREAM_FOOTER_SIZE) + return SZ_ERROR_NO_ARCHIVE; + + pos -= XZ_STREAM_FOOTER_SIZE; + RINOK(LookInStream_SeekRead_ForArc(stream, pos, buf, XZ_STREAM_FOOTER_SIZE)); + + if (!XZ_FOOTER_SIG_CHECK(buf + 10)) + { + UInt32 total = 0; + pos += XZ_STREAM_FOOTER_SIZE; + + for (;;) + { + size_t i; + #define TEMP_BUF_SIZE (1 << 10) + Byte temp[TEMP_BUF_SIZE]; + + i = (pos > TEMP_BUF_SIZE) ? TEMP_BUF_SIZE : (size_t)pos; + pos -= i; + RINOK(LookInStream_SeekRead_ForArc(stream, pos, temp, i)); + total += (UInt32)i; + for (; i != 0; i--) + if (temp[i - 1] != 0) + break; + if (i != 0) + { + if ((i & 3) != 0) + return SZ_ERROR_NO_ARCHIVE; + pos += i; + break; + } + if (pos < XZ_STREAM_FOOTER_SIZE || total > (1 << 16)) + return SZ_ERROR_NO_ARCHIVE; + } + + if (pos < XZ_STREAM_FOOTER_SIZE) + return SZ_ERROR_NO_ARCHIVE; + pos -= XZ_STREAM_FOOTER_SIZE; + RINOK(LookInStream_SeekRead_ForArc(stream, pos, buf, XZ_STREAM_FOOTER_SIZE)); + if (!XZ_FOOTER_SIG_CHECK(buf + 10)) + return SZ_ERROR_NO_ARCHIVE; + } + + p->flags = (CXzStreamFlags)GetBe16(buf + 8); + + if (!XzFlags_IsSupported(p->flags)) + return SZ_ERROR_UNSUPPORTED; + + { + /* to eliminate GCC 6.3 warning: + dereferencing type-punned pointer will break strict-aliasing rules */ + const Byte *buf_ptr = buf; + if (GetUi32(buf_ptr) != CrcCalc(buf + 4, 6)) + return SZ_ERROR_ARCHIVE; + } + + indexSize = ((UInt64)GetUi32(buf + 4) + 1) << 2; + + if (pos < indexSize) + return SZ_ERROR_ARCHIVE; + + pos -= indexSize; + RINOK(LookInStream_SeekTo(stream, pos)); + RINOK(Xz_ReadIndex(p, stream, indexSize, alloc)); + + { + UInt64 totalSize = Xz_GetPackSize(p); + if (totalSize == XZ_SIZE_OVERFLOW + || totalSize >= ((UInt64)1 << 63) + || pos < totalSize + XZ_STREAM_HEADER_SIZE) + return SZ_ERROR_ARCHIVE; + pos -= (totalSize + XZ_STREAM_HEADER_SIZE); + RINOK(LookInStream_SeekTo(stream, pos)); + *startOffset = (Int64)pos; + } + { + CXzStreamFlags headerFlags; + CSecToRead secToRead; + SecToRead_CreateVTable(&secToRead); + secToRead.realStream = stream; + + RINOK(Xz_ReadHeader(&headerFlags, &secToRead.vt)); + return (p->flags == headerFlags) ? SZ_OK : SZ_ERROR_ARCHIVE; + } +} + + +/* ---------- Xz Streams ---------- */ + +void Xzs_Construct(CXzs *p) +{ + p->num = p->numAllocated = 0; + p->streams = 0; +} + +void Xzs_Free(CXzs *p, ISzAllocPtr alloc) +{ + size_t i; + for (i = 0; i < p->num; i++) + Xz_Free(&p->streams[i], alloc); + ISzAlloc_Free(alloc, p->streams); + p->num = p->numAllocated = 0; + p->streams = 0; +} + +UInt64 Xzs_GetNumBlocks(const CXzs *p) +{ + UInt64 num = 0; + size_t i; + for (i = 0; i < p->num; i++) + num += p->streams[i].numBlocks; + return num; +} + +UInt64 Xzs_GetUnpackSize(const CXzs *p) +{ + UInt64 size = 0; + size_t i; + for (i = 0; i < p->num; i++) + ADD_SIZE_CHECK(size, Xz_GetUnpackSize(&p->streams[i])); + return size; +} + +/* +UInt64 Xzs_GetPackSize(const CXzs *p) +{ + UInt64 size = 0; + size_t i; + for (i = 0; i < p->num; i++) + ADD_SIZE_CHECK(size, Xz_GetTotalSize(&p->streams[i])); + return size; +} +*/ + +SRes Xzs_ReadBackward(CXzs *p, ILookInStream *stream, Int64 *startOffset, ICompressProgress *progress, ISzAllocPtr alloc) +{ + Int64 endOffset = 0; + RINOK(ILookInStream_Seek(stream, &endOffset, SZ_SEEK_END)); + *startOffset = endOffset; + for (;;) + { + CXzStream st; + SRes res; + Xz_Construct(&st); + res = Xz_ReadBackward(&st, stream, startOffset, alloc); + st.startOffset = (UInt64)*startOffset; + RINOK(res); + if (p->num == p->numAllocated) + { + const size_t newNum = p->num + p->num / 4 + 1; + void *data = ISzAlloc_Alloc(alloc, newNum * sizeof(CXzStream)); + if (!data) + return SZ_ERROR_MEM; + p->numAllocated = newNum; + if (p->num != 0) + memcpy(data, p->streams, p->num * sizeof(CXzStream)); + ISzAlloc_Free(alloc, p->streams); + p->streams = (CXzStream *)data; + } + p->streams[p->num++] = st; + if (*startOffset == 0) + break; + RINOK(LookInStream_SeekTo(stream, (UInt64)*startOffset)); + if (progress && ICompressProgress_Progress(progress, (UInt64)(endOffset - *startOffset), (UInt64)(Int64)-1) != SZ_OK) + return SZ_ERROR_PROGRESS; + } + return SZ_OK; +} diff --git a/deps/LZMA-SDK/C/var_clang.mak b/deps/LZMA-SDK/C/var_clang.mak index ee265698e..a6df26e77 100644 --- a/deps/LZMA-SDK/C/var_clang.mak +++ b/deps/LZMA-SDK/C/var_clang.mak @@ -1,11 +1,11 @@ -PLATFORM= -O=b/c -IS_X64= -IS_X86= -IS_ARM64= -CROSS_COMPILE= -MY_ARCH= -USE_ASM= -CC=$(CROSS_COMPILE)clang -CXX=$(CROSS_COMPILE)clang++ -USE_CLANG=1 +PLATFORM= +O=b/c +IS_X64= +IS_X86= +IS_ARM64= +CROSS_COMPILE= +MY_ARCH= +USE_ASM= +CC=$(CROSS_COMPILE)clang +CXX=$(CROSS_COMPILE)clang++ +USE_CLANG=1 diff --git a/deps/LZMA-SDK/C/var_clang_arm64.mak b/deps/LZMA-SDK/C/var_clang_arm64.mak index 1e82d2eb6..4b3540969 100644 --- a/deps/LZMA-SDK/C/var_clang_arm64.mak +++ b/deps/LZMA-SDK/C/var_clang_arm64.mak @@ -1,11 +1,11 @@ -PLATFORM=arm64 -O=b/c_$(PLATFORM) -IS_X64= -IS_X86= -IS_ARM64=1 -CROSS_COMPILE= -MY_ARCH= -USE_ASM=1 -CC=$(CROSS_COMPILE)clang -CXX=$(CROSS_COMPILE)clang++ -USE_CLANG=1 +PLATFORM=arm64 +O=b/c_$(PLATFORM) +IS_X64= +IS_X86= +IS_ARM64=1 +CROSS_COMPILE= +MY_ARCH= +USE_ASM=1 +CC=$(CROSS_COMPILE)clang +CXX=$(CROSS_COMPILE)clang++ +USE_CLANG=1 diff --git a/deps/LZMA-SDK/C/var_clang_x64.mak b/deps/LZMA-SDK/C/var_clang_x64.mak index b35f0cca4..34e1b49c5 100644 --- a/deps/LZMA-SDK/C/var_clang_x64.mak +++ b/deps/LZMA-SDK/C/var_clang_x64.mak @@ -1,11 +1,11 @@ -PLATFORM=x64 -O=b/c_$(PLATFORM) -IS_X64=1 -IS_X86= -IS_ARM64= -CROSS_COMPILE= -MY_ARCH= -USE_ASM=1 -CC=$(CROSS_COMPILE)clang -CXX=$(CROSS_COMPILE)clang++ -USE_CLANG=1 +PLATFORM=x64 +O=b/c_$(PLATFORM) +IS_X64=1 +IS_X86= +IS_ARM64= +CROSS_COMPILE= +MY_ARCH= +USE_ASM=1 +CC=$(CROSS_COMPILE)clang +CXX=$(CROSS_COMPILE)clang++ +USE_CLANG=1 diff --git a/deps/LZMA-SDK/C/var_clang_x86.mak b/deps/LZMA-SDK/C/var_clang_x86.mak index d18f0ae12..bd2317c29 100644 --- a/deps/LZMA-SDK/C/var_clang_x86.mak +++ b/deps/LZMA-SDK/C/var_clang_x86.mak @@ -1,11 +1,11 @@ -PLATFORM=x86 -O=b/c_$(PLATFORM) -IS_X64= -IS_X86=1 -IS_ARM64= -CROSS_COMPILE= -MY_ARCH=-m32 -USE_ASM=1 -CC=$(CROSS_COMPILE)clang -CXX=$(CROSS_COMPILE)clang++ -USE_CLANG=1 +PLATFORM=x86 +O=b/c_$(PLATFORM) +IS_X64= +IS_X86=1 +IS_ARM64= +CROSS_COMPILE= +MY_ARCH=-m32 +USE_ASM=1 +CC=$(CROSS_COMPILE)clang +CXX=$(CROSS_COMPILE)clang++ +USE_CLANG=1 diff --git a/deps/LZMA-SDK/C/var_gcc.mak b/deps/LZMA-SDK/C/var_gcc.mak index 803c8de9d..664491cf2 100644 --- a/deps/LZMA-SDK/C/var_gcc.mak +++ b/deps/LZMA-SDK/C/var_gcc.mak @@ -1,12 +1,12 @@ -PLATFORM= -O=b/g -IS_X64= -IS_X86= -IS_ARM64= -CROSS_COMPILE= -MY_ARCH= -USE_ASM= -CC=$(CROSS_COMPILE)gcc -CXX=$(CROSS_COMPILE)g++ - -# -march=armv8-a+crc+crypto +PLATFORM= +O=b/g +IS_X64= +IS_X86= +IS_ARM64= +CROSS_COMPILE= +MY_ARCH= +USE_ASM= +CC=$(CROSS_COMPILE)gcc +CXX=$(CROSS_COMPILE)g++ + +# -march=armv8-a+crc+crypto diff --git a/deps/LZMA-SDK/C/var_gcc_arm64.mak b/deps/LZMA-SDK/C/var_gcc_arm64.mak index 562cfaa6d..4bbb687db 100644 --- a/deps/LZMA-SDK/C/var_gcc_arm64.mak +++ b/deps/LZMA-SDK/C/var_gcc_arm64.mak @@ -1,12 +1,12 @@ -PLATFORM=arm64 -O=b/g_$(PLATFORM) -IS_X64= -IS_X86= -IS_ARM64=1 -CROSS_COMPILE= -MY_ARCH=-mtune=cortex-a53 -USE_ASM=1 -CC=$(CROSS_COMPILE)gcc -CXX=$(CROSS_COMPILE)g++ - -# -march=armv8-a+crc+crypto +PLATFORM=arm64 +O=b/g_$(PLATFORM) +IS_X64= +IS_X86= +IS_ARM64=1 +CROSS_COMPILE= +MY_ARCH=-mtune=cortex-a53 +USE_ASM=1 +CC=$(CROSS_COMPILE)gcc +CXX=$(CROSS_COMPILE)g++ + +# -march=armv8-a+crc+crypto diff --git a/deps/LZMA-SDK/C/var_gcc_x64.mak b/deps/LZMA-SDK/C/var_gcc_x64.mak index 1b965b21b..1acf604f5 100644 --- a/deps/LZMA-SDK/C/var_gcc_x64.mak +++ b/deps/LZMA-SDK/C/var_gcc_x64.mak @@ -1,10 +1,10 @@ -PLATFORM=x64 -O=b/g_$(PLATFORM) -IS_X64=1 -IS_X86= -IS_ARM64= -CROSS_COMPILE= -MY_ARCH= -USE_ASM=1 -CC=$(CROSS_COMPILE)gcc -CXX=$(CROSS_COMPILE)g++ +PLATFORM=x64 +O=b/g_$(PLATFORM) +IS_X64=1 +IS_X86= +IS_ARM64= +CROSS_COMPILE= +MY_ARCH= +USE_ASM=1 +CC=$(CROSS_COMPILE)gcc +CXX=$(CROSS_COMPILE)g++ diff --git a/deps/LZMA-SDK/C/var_gcc_x86.mak b/deps/LZMA-SDK/C/var_gcc_x86.mak index 3adf4a84a..f0718ec71 100644 --- a/deps/LZMA-SDK/C/var_gcc_x86.mak +++ b/deps/LZMA-SDK/C/var_gcc_x86.mak @@ -1,10 +1,10 @@ -PLATFORM=x86 -O=b/g_$(PLATFORM) -IS_X64= -IS_X86=1 -IS_ARM64= -CROSS_COMPILE= -MY_ARCH=-m32 -USE_ASM=1 -CC=$(CROSS_COMPILE)gcc -CXX=$(CROSS_COMPILE)g++ +PLATFORM=x86 +O=b/g_$(PLATFORM) +IS_X64= +IS_X86=1 +IS_ARM64= +CROSS_COMPILE= +MY_ARCH=-m32 +USE_ASM=1 +CC=$(CROSS_COMPILE)gcc +CXX=$(CROSS_COMPILE)g++ diff --git a/deps/LZMA-SDK/C/var_mac_arm64.mak b/deps/LZMA-SDK/C/var_mac_arm64.mak index 0ba414230..adf5fa1d7 100644 --- a/deps/LZMA-SDK/C/var_mac_arm64.mak +++ b/deps/LZMA-SDK/C/var_mac_arm64.mak @@ -1,11 +1,11 @@ -PLATFORM=arm64 -O=b/m_$(PLATFORM) -IS_X64= -IS_X86= -IS_ARM64=1 -CROSS_COMPILE= -MY_ARCH=-arch arm64 -USE_ASM=1 -CC=$(CROSS_COMPILE)clang -CXX=$(CROSS_COMPILE)clang++ -USE_CLANG=1 +PLATFORM=arm64 +O=b/m_$(PLATFORM) +IS_X64= +IS_X86= +IS_ARM64=1 +CROSS_COMPILE= +MY_ARCH=-arch arm64 +USE_ASM=1 +CC=$(CROSS_COMPILE)clang +CXX=$(CROSS_COMPILE)clang++ +USE_CLANG=1 diff --git a/deps/LZMA-SDK/C/var_mac_x64.mak b/deps/LZMA-SDK/C/var_mac_x64.mak index 92b15c8b7..13d7aa7fe 100644 --- a/deps/LZMA-SDK/C/var_mac_x64.mak +++ b/deps/LZMA-SDK/C/var_mac_x64.mak @@ -1,11 +1,11 @@ -PLATFORM=x64 -O=b/m_$(PLATFORM) -IS_X64=1 -IS_X86= -IS_ARM64= -CROSS_COMPILE= -MY_ARCH=-arch x86_64 -USE_ASM= -CC=$(CROSS_COMPILE)clang -CXX=$(CROSS_COMPILE)clang++ -USE_CLANG=1 +PLATFORM=x64 +O=b/m_$(PLATFORM) +IS_X64=1 +IS_X86= +IS_ARM64= +CROSS_COMPILE= +MY_ARCH=-arch x86_64 +USE_ASM= +CC=$(CROSS_COMPILE)clang +CXX=$(CROSS_COMPILE)clang++ +USE_CLANG=1 diff --git a/deps/LZMA-SDK/C/warn_clang.mak b/deps/LZMA-SDK/C/warn_clang.mak index a299fbc4d..ed4f908f6 100644 --- a/deps/LZMA-SDK/C/warn_clang.mak +++ b/deps/LZMA-SDK/C/warn_clang.mak @@ -1,37 +1,37 @@ -CFLAGS_WARN_CLANG_3_8_UNIQ = \ - -Wno-reserved-id-macro \ - -Wno-old-style-cast \ - -Wno-c++11-long-long \ - -Wno-unused-macros \ - -CFLAGS_WARN_CLANG_3_8 = \ - $(CFLAGS_WARN_CLANG_3_8_UNIQ) \ - -Weverything \ - -Wno-extra-semi \ - -Wno-sign-conversion \ - -Wno-language-extension-token \ - -Wno-global-constructors \ - -Wno-non-virtual-dtor \ - -Wno-switch-enum \ - -Wno-covered-switch-default \ - -Wno-cast-qual \ - -Wno-padded \ - -Wno-exit-time-destructors \ - -Wno-weak-vtables \ - -CFLAGS_WARN_CLANG_12= $(CFLAGS_WARN_CLANG_3_8) \ - -Wno-extra-semi-stmt \ - -Wno-zero-as-null-pointer-constant \ - -Wno-deprecated-dynamic-exception-spec \ - -Wno-c++98-compat-pedantic \ - -Wno-atomic-implicit-seq-cst \ - -Wconversion \ - -Wno-sign-conversion \ - -CFLAGS_WARN_1 = \ - -Wno-deprecated-copy-dtor \ - - - - -CFLAGS_WARN = $(CFLAGS_WARN_CLANG_12) $(CFLAGS_WARN_1) +CFLAGS_WARN_CLANG_3_8_UNIQ = \ + -Wno-reserved-id-macro \ + -Wno-old-style-cast \ + -Wno-c++11-long-long \ + -Wno-unused-macros \ + +CFLAGS_WARN_CLANG_3_8 = \ + $(CFLAGS_WARN_CLANG_3_8_UNIQ) \ + -Weverything \ + -Wno-extra-semi \ + -Wno-sign-conversion \ + -Wno-language-extension-token \ + -Wno-global-constructors \ + -Wno-non-virtual-dtor \ + -Wno-switch-enum \ + -Wno-covered-switch-default \ + -Wno-cast-qual \ + -Wno-padded \ + -Wno-exit-time-destructors \ + -Wno-weak-vtables \ + +CFLAGS_WARN_CLANG_12= $(CFLAGS_WARN_CLANG_3_8) \ + -Wno-extra-semi-stmt \ + -Wno-zero-as-null-pointer-constant \ + -Wno-deprecated-dynamic-exception-spec \ + -Wno-c++98-compat-pedantic \ + -Wno-atomic-implicit-seq-cst \ + -Wconversion \ + -Wno-sign-conversion \ + +CFLAGS_WARN_1 = \ + -Wno-deprecated-copy-dtor \ + + + + +CFLAGS_WARN = $(CFLAGS_WARN_CLANG_12) $(CFLAGS_WARN_1) diff --git a/deps/LZMA-SDK/C/warn_clang_mac.mak b/deps/LZMA-SDK/C/warn_clang_mac.mak index cfbbda073..41044a2ce 100644 --- a/deps/LZMA-SDK/C/warn_clang_mac.mak +++ b/deps/LZMA-SDK/C/warn_clang_mac.mak @@ -1,37 +1,37 @@ -CFLAGS_WARN_CLANG_3_8_UNIQ = \ - -Wno-reserved-id-macro \ - -Wno-old-style-cast \ - -Wno-c++11-long-long \ - -Wno-unused-macros \ - -CFLAGS_WARN_CLANG_3_8 = \ - $(CFLAGS_WARN_CLANG_3_8_UNIQ) \ - -Weverything \ - -Wno-extra-semi \ - -Wno-sign-conversion \ - -Wno-language-extension-token \ - -Wno-global-constructors \ - -Wno-non-virtual-dtor \ - -Wno-switch-enum \ - -Wno-covered-switch-default \ - -Wno-cast-qual \ - -Wno-padded \ - -Wno-exit-time-destructors \ - -Wno-weak-vtables \ - -CFLAGS_WARN_CLANG_12= $(CFLAGS_WARN_CLANG_3_8) \ - -Wno-extra-semi-stmt \ - -Wno-zero-as-null-pointer-constant \ - -Wno-deprecated-dynamic-exception-spec \ - -Wno-c++98-compat-pedantic \ - -Wno-atomic-implicit-seq-cst \ - -Wconversion \ - -Wno-sign-conversion \ - -CFLAGS_WARN_MAC = \ - -Wno-poison-system-directories \ - -Wno-c++11-long-long \ - -Wno-atomic-implicit-seq-cst \ - - -CFLAGS_WARN = $(CFLAGS_WARN_CLANG_12) $(CFLAGS_WARN_MAC) +CFLAGS_WARN_CLANG_3_8_UNIQ = \ + -Wno-reserved-id-macro \ + -Wno-old-style-cast \ + -Wno-c++11-long-long \ + -Wno-unused-macros \ + +CFLAGS_WARN_CLANG_3_8 = \ + $(CFLAGS_WARN_CLANG_3_8_UNIQ) \ + -Weverything \ + -Wno-extra-semi \ + -Wno-sign-conversion \ + -Wno-language-extension-token \ + -Wno-global-constructors \ + -Wno-non-virtual-dtor \ + -Wno-switch-enum \ + -Wno-covered-switch-default \ + -Wno-cast-qual \ + -Wno-padded \ + -Wno-exit-time-destructors \ + -Wno-weak-vtables \ + +CFLAGS_WARN_CLANG_12= $(CFLAGS_WARN_CLANG_3_8) \ + -Wno-extra-semi-stmt \ + -Wno-zero-as-null-pointer-constant \ + -Wno-deprecated-dynamic-exception-spec \ + -Wno-c++98-compat-pedantic \ + -Wno-atomic-implicit-seq-cst \ + -Wconversion \ + -Wno-sign-conversion \ + +CFLAGS_WARN_MAC = \ + -Wno-poison-system-directories \ + -Wno-c++11-long-long \ + -Wno-atomic-implicit-seq-cst \ + + +CFLAGS_WARN = $(CFLAGS_WARN_CLANG_12) $(CFLAGS_WARN_MAC) diff --git a/deps/LZMA-SDK/C/warn_gcc.mak b/deps/LZMA-SDK/C/warn_gcc.mak index 8fbdb28f6..7aab7a449 100644 --- a/deps/LZMA-SDK/C/warn_gcc.mak +++ b/deps/LZMA-SDK/C/warn_gcc.mak @@ -1,51 +1,51 @@ -CFLAGS_WARN_GCC_4_5 = \ - -CFLAGS_WARN_GCC_6 = \ - -Waddress \ - -Waggressive-loop-optimizations \ - -Wattributes \ - -Wbool-compare \ - -Wcast-align \ - -Wcomment \ - -Wdiv-by-zero \ - -Wduplicated-cond \ - -Wformat-contains-nul \ - -Winit-self \ - -Wint-to-pointer-cast \ - -Wunused \ - -Wunused-macros \ - -# -Wno-strict-aliasing - -CFLAGS_WARN_GCC_9 = \ - -Waddress \ - -Waddress-of-packed-member \ - -Waggressive-loop-optimizations \ - -Wattributes \ - -Wbool-compare \ - -Wbool-operation \ - -Wcast-align \ - -Wcast-align=strict \ - -Wcomment \ - -Wdangling-else \ - -Wdiv-by-zero \ - -Wduplicated-branches \ - -Wduplicated-cond \ - -Wformat-contains-nul \ - -Wimplicit-fallthrough=5 \ - -Winit-self \ - -Wint-in-bool-context \ - -Wint-to-pointer-cast \ - -Wunused \ - -Wunused-macros \ - -Wconversion \ - -# -Wno-sign-conversion \ - -CFLAGS_WARN_GCC_PPMD_UNALIGNED = \ - -Wno-strict-aliasing \ - - -CFLAGS_WARN = $(CFLAGS_WARN_GCC_9) \ - -# $(CFLAGS_WARN_GCC_PPMD_UNALIGNED) +CFLAGS_WARN_GCC_4_5 = \ + +CFLAGS_WARN_GCC_6 = \ + -Waddress \ + -Waggressive-loop-optimizations \ + -Wattributes \ + -Wbool-compare \ + -Wcast-align \ + -Wcomment \ + -Wdiv-by-zero \ + -Wduplicated-cond \ + -Wformat-contains-nul \ + -Winit-self \ + -Wint-to-pointer-cast \ + -Wunused \ + -Wunused-macros \ + +# -Wno-strict-aliasing + +CFLAGS_WARN_GCC_9 = \ + -Waddress \ + -Waddress-of-packed-member \ + -Waggressive-loop-optimizations \ + -Wattributes \ + -Wbool-compare \ + -Wbool-operation \ + -Wcast-align \ + -Wcast-align=strict \ + -Wcomment \ + -Wdangling-else \ + -Wdiv-by-zero \ + -Wduplicated-branches \ + -Wduplicated-cond \ + -Wformat-contains-nul \ + -Wimplicit-fallthrough=5 \ + -Winit-self \ + -Wint-in-bool-context \ + -Wint-to-pointer-cast \ + -Wunused \ + -Wunused-macros \ + -Wconversion \ + +# -Wno-sign-conversion \ + +CFLAGS_WARN_GCC_PPMD_UNALIGNED = \ + -Wno-strict-aliasing \ + + +CFLAGS_WARN = $(CFLAGS_WARN_GCC_9) \ + +# $(CFLAGS_WARN_GCC_PPMD_UNALIGNED) diff --git a/deps/LZMA-SDK/DOC/7zC.txt b/deps/LZMA-SDK/DOC/7zC.txt index 49276787f..939b720f9 100644 --- a/deps/LZMA-SDK/DOC/7zC.txt +++ b/deps/LZMA-SDK/DOC/7zC.txt @@ -1,187 +1,187 @@ -7z ANSI-C Decoder 9.35 ----------------------- - -7z ANSI-C provides 7z/LZMA decoding. -7z ANSI-C version is simplified version ported from C++ code. - -LZMA is default and general compression method of 7z format -in 7-Zip compression program (www.7-zip.org). LZMA provides high -compression ratio and very fast decompression. - - -LICENSE -------- - -7z ANSI-C Decoder is part of the LZMA SDK. -LZMA SDK is written and placed in the public domain by Igor Pavlov. - -Files ---------------------- - -7zDecode.* - Low level 7z decoding -7zExtract.* - High level 7z decoding -7zHeader.* - .7z format constants -7zIn.* - .7z archive opening -7zItem.* - .7z structures -7zMain.c - Test application - - -How To Use ----------- - -You can create .7z archive with 7z.exe, 7za.exe or 7zr.exe: - - 7z.exe a archive.7z *.htm -r -mx -m0fb=255 - -If you have big number of files in archive, and you need fast extracting, -you can use partly-solid archives: - - 7za.exe a archive.7z *.htm -ms=512K -r -mx -m0fb=255 -m0d=512K - -In that example 7-Zip will use 512KB solid blocks. So it needs to decompress only -512KB for extracting one file from such archive. - - -Limitations of current version of 7z ANSI-C Decoder ---------------------------------------------------- - - - It reads only "FileName", "Size", "LastWriteTime" and "CRC" information for each file in archive. - - It supports only LZMA and Copy (no compression) methods with BCJ or BCJ2 filters. - - It converts original UTF-16 Unicode file names to UTF-8 Unicode file names. - -These limitations will be fixed in future versions. - - -Using 7z ANSI-C Decoder Test application: ------------------------------------------ - -Usage: 7zDec - -: - e: Extract files from archive - l: List contents of archive - t: Test integrity of archive - -Example: - - 7zDec l archive.7z - -lists contents of archive.7z - - 7zDec e archive.7z - -extracts files from archive.7z to current folder. - - -How to use .7z Decoder ----------------------- - -Memory allocation -~~~~~~~~~~~~~~~~~ - -7z Decoder uses two memory pools: -1) Temporary pool -2) Main pool -Such scheme can allow you to avoid fragmentation of allocated blocks. - - -Steps for using 7z decoder --------------------------- - -Use code at 7zMain.c as example. - -1) Declare variables: - inStream /* implements ILookInStream interface */ - CSzArEx db; /* 7z archive database structure */ - ISzAlloc allocImp; /* memory functions for main pool */ - ISzAlloc allocTempImp; /* memory functions for temporary pool */ - -2) call CrcGenerateTable(); function to initialize CRC structures. - -3) call SzArEx_Init(&db); function to initialize db structures. - -4) call SzArEx_Open(&db, inStream, &allocMain, &allocTemp) to open archive - -This function opens archive "inStream" and reads headers to "db". -All items in "db" will be allocated with "allocMain" functions. -SzArEx_Open function allocates and frees temporary structures by "allocTemp" functions. - -5) List items or Extract items - - Listing code: - ~~~~~~~~~~~~~ - - Use SzArEx_GetFileNameUtf16 function. Look example code in C\Util\7z\7zMain.c file. - - - Extracting code: - ~~~~~~~~~~~~~~~~ - - SZ_RESULT SzAr_Extract( - CArchiveDatabaseEx *db, - ILookInStream *inStream, - UInt32 fileIndex, /* index of file */ - UInt32 *blockIndex, /* index of solid block */ - Byte **outBuffer, /* pointer to pointer to output buffer (allocated with allocMain) */ - size_t *outBufferSize, /* buffer size for output buffer */ - size_t *offset, /* offset of stream for required file in *outBuffer */ - size_t *outSizeProcessed, /* size of file in *outBuffer */ - ISzAlloc *allocMain, - ISzAlloc *allocTemp); - - If you need to decompress more than one file, you can send these values from previous call: - blockIndex, - outBuffer, - outBufferSize, - You can consider "outBuffer" as cache of solid block. If your archive is solid, - it will increase decompression speed. - - After decompressing you must free "outBuffer": - allocImp.Free(outBuffer); - -6) call SzArEx_Free(&db, allocImp.Free) to free allocated items in "db". - - - - -Memory requirements for .7z decoding ------------------------------------- - -Memory usage for Archive opening: - - Temporary pool: - - Memory for uncompressed .7z headers - - some other temporary blocks - - Main pool: - - Memory for database: - Estimated size of one file structures in solid archive: - - Size (4 or 8 Bytes) - - CRC32 (4 bytes) - - LastWriteTime (8 bytes) - - Some file information (4 bytes) - - File Name (variable length) + pointer + allocation structures - -Memory usage for archive Decompressing: - - Temporary pool: - - Memory for LZMA decompressing structures - - Main pool: - - Memory for decompressed solid block - - Memory for temprorary buffers, if BCJ2 fileter is used. Usually these - temprorary buffers can be about 15% of solid block size. - - -7z Decoder doesn't allocate memory for compressed blocks. -Instead of this, you must allocate buffer with desired -size before calling 7z Decoder. Use 7zMain.c as example. - - -Defines -------- - -_SZ_ALLOC_DEBUG - define it if you want to debug alloc/free operations to stderr. - - ---- - -http://www.7-zip.org -http://www.7-zip.org/sdk.html -http://www.7-zip.org/support.html +7z ANSI-C Decoder 9.35 +---------------------- + +7z ANSI-C provides 7z/LZMA decoding. +7z ANSI-C version is simplified version ported from C++ code. + +LZMA is default and general compression method of 7z format +in 7-Zip compression program (www.7-zip.org). LZMA provides high +compression ratio and very fast decompression. + + +LICENSE +------- + +7z ANSI-C Decoder is part of the LZMA SDK. +LZMA SDK is written and placed in the public domain by Igor Pavlov. + +Files +--------------------- + +7zDecode.* - Low level 7z decoding +7zExtract.* - High level 7z decoding +7zHeader.* - .7z format constants +7zIn.* - .7z archive opening +7zItem.* - .7z structures +7zMain.c - Test application + + +How To Use +---------- + +You can create .7z archive with 7z.exe, 7za.exe or 7zr.exe: + + 7z.exe a archive.7z *.htm -r -mx -m0fb=255 + +If you have big number of files in archive, and you need fast extracting, +you can use partly-solid archives: + + 7za.exe a archive.7z *.htm -ms=512K -r -mx -m0fb=255 -m0d=512K + +In that example 7-Zip will use 512KB solid blocks. So it needs to decompress only +512KB for extracting one file from such archive. + + +Limitations of current version of 7z ANSI-C Decoder +--------------------------------------------------- + + - It reads only "FileName", "Size", "LastWriteTime" and "CRC" information for each file in archive. + - It supports only LZMA and Copy (no compression) methods with BCJ or BCJ2 filters. + - It converts original UTF-16 Unicode file names to UTF-8 Unicode file names. + +These limitations will be fixed in future versions. + + +Using 7z ANSI-C Decoder Test application: +----------------------------------------- + +Usage: 7zDec + +: + e: Extract files from archive + l: List contents of archive + t: Test integrity of archive + +Example: + + 7zDec l archive.7z + +lists contents of archive.7z + + 7zDec e archive.7z + +extracts files from archive.7z to current folder. + + +How to use .7z Decoder +---------------------- + +Memory allocation +~~~~~~~~~~~~~~~~~ + +7z Decoder uses two memory pools: +1) Temporary pool +2) Main pool +Such scheme can allow you to avoid fragmentation of allocated blocks. + + +Steps for using 7z decoder +-------------------------- + +Use code at 7zMain.c as example. + +1) Declare variables: + inStream /* implements ILookInStream interface */ + CSzArEx db; /* 7z archive database structure */ + ISzAlloc allocImp; /* memory functions for main pool */ + ISzAlloc allocTempImp; /* memory functions for temporary pool */ + +2) call CrcGenerateTable(); function to initialize CRC structures. + +3) call SzArEx_Init(&db); function to initialize db structures. + +4) call SzArEx_Open(&db, inStream, &allocMain, &allocTemp) to open archive + +This function opens archive "inStream" and reads headers to "db". +All items in "db" will be allocated with "allocMain" functions. +SzArEx_Open function allocates and frees temporary structures by "allocTemp" functions. + +5) List items or Extract items + + Listing code: + ~~~~~~~~~~~~~ + + Use SzArEx_GetFileNameUtf16 function. Look example code in C\Util\7z\7zMain.c file. + + + Extracting code: + ~~~~~~~~~~~~~~~~ + + SZ_RESULT SzAr_Extract( + CArchiveDatabaseEx *db, + ILookInStream *inStream, + UInt32 fileIndex, /* index of file */ + UInt32 *blockIndex, /* index of solid block */ + Byte **outBuffer, /* pointer to pointer to output buffer (allocated with allocMain) */ + size_t *outBufferSize, /* buffer size for output buffer */ + size_t *offset, /* offset of stream for required file in *outBuffer */ + size_t *outSizeProcessed, /* size of file in *outBuffer */ + ISzAlloc *allocMain, + ISzAlloc *allocTemp); + + If you need to decompress more than one file, you can send these values from previous call: + blockIndex, + outBuffer, + outBufferSize, + You can consider "outBuffer" as cache of solid block. If your archive is solid, + it will increase decompression speed. + + After decompressing you must free "outBuffer": + allocImp.Free(outBuffer); + +6) call SzArEx_Free(&db, allocImp.Free) to free allocated items in "db". + + + + +Memory requirements for .7z decoding +------------------------------------ + +Memory usage for Archive opening: + - Temporary pool: + - Memory for uncompressed .7z headers + - some other temporary blocks + - Main pool: + - Memory for database: + Estimated size of one file structures in solid archive: + - Size (4 or 8 Bytes) + - CRC32 (4 bytes) + - LastWriteTime (8 bytes) + - Some file information (4 bytes) + - File Name (variable length) + pointer + allocation structures + +Memory usage for archive Decompressing: + - Temporary pool: + - Memory for LZMA decompressing structures + - Main pool: + - Memory for decompressed solid block + - Memory for temprorary buffers, if BCJ2 fileter is used. Usually these + temprorary buffers can be about 15% of solid block size. + + +7z Decoder doesn't allocate memory for compressed blocks. +Instead of this, you must allocate buffer with desired +size before calling 7z Decoder. Use 7zMain.c as example. + + +Defines +------- + +_SZ_ALLOC_DEBUG - define it if you want to debug alloc/free operations to stderr. + + +--- + +http://www.7-zip.org +http://www.7-zip.org/sdk.html +http://www.7-zip.org/support.html diff --git a/deps/LZMA-SDK/DOC/7zFormat.txt b/deps/LZMA-SDK/DOC/7zFormat.txt index 9239e9355..74cdfa418 100644 --- a/deps/LZMA-SDK/DOC/7zFormat.txt +++ b/deps/LZMA-SDK/DOC/7zFormat.txt @@ -1,469 +1,469 @@ -7z Format description (18.06) ----------------------------- - -This file contains description of 7z archive format. -7z archive can contain files compressed with any method. -See "Methods.txt" for description for defined compressing methods. - - -Format structure Overview -------------------------- - -Some fields can be optional. - -Archive structure -~~~~~~~~~~~~~~~~~ -SignatureHeader -[PackedStreams] -[PackedStreamsForHeaders] -[ - Header - or - { - Packed Header - HeaderInfo - } -] - - - -Header structure -~~~~~~~~~~~~~~~~ -{ - ArchiveProperties - AdditionalStreams - { - PackInfo - { - PackPos - NumPackStreams - Sizes[NumPackStreams] - CRCs[NumPackStreams] - } - CodersInfo - { - NumFolders - Folders[NumFolders] - { - NumCoders - CodersInfo[NumCoders] - { - ID - NumInStreams; - NumOutStreams; - PropertiesSize - Properties[PropertiesSize] - } - NumBindPairs - BindPairsInfo[NumBindPairs] - { - InIndex; - OutIndex; - } - PackedIndices - } - UnPackSize[Folders][Folders.NumOutstreams] - CRCs[NumFolders] - } - SubStreamsInfo - { - NumUnPackStreamsInFolders[NumFolders]; - UnPackSizes[] - CRCs[] - } - } - MainStreamsInfo - { - (Same as in AdditionalStreams) - } - FilesInfo - { - NumFiles - Properties[] - { - ID - Size - Data - } - } -} - -HeaderInfo structure -~~~~~~~~~~~~~~~~~~~~ -{ - (Same as in AdditionalStreams) -} - - - -Notes about Notation and encoding ---------------------------------- - -7z uses little endian encoding. - -7z archive format has optional headers that are marked as -[] -Header -[] - -REAL_UINT64 means real UINT64. - -UINT64 means real UINT64 encoded with the following scheme: - - Size of encoding sequence depends from first byte: - First_Byte Extra_Bytes Value - (binary) - 0xxxxxxx : ( xxxxxxx ) - 10xxxxxx BYTE y[1] : ( xxxxxx << (8 * 1)) + y - 110xxxxx BYTE y[2] : ( xxxxx << (8 * 2)) + y - ... - 1111110x BYTE y[6] : ( x << (8 * 6)) + y - 11111110 BYTE y[7] : y - 11111111 BYTE y[8] : y - - - -Property IDs ------------- - -0x00 = kEnd - -0x01 = kHeader - -0x02 = kArchiveProperties - -0x03 = kAdditionalStreamsInfo -0x04 = kMainStreamsInfo -0x05 = kFilesInfo - -0x06 = kPackInfo -0x07 = kUnPackInfo -0x08 = kSubStreamsInfo - -0x09 = kSize -0x0A = kCRC - -0x0B = kFolder - -0x0C = kCodersUnPackSize -0x0D = kNumUnPackStream - -0x0E = kEmptyStream -0x0F = kEmptyFile -0x10 = kAnti - -0x11 = kName -0x12 = kCTime -0x13 = kATime -0x14 = kMTime -0x15 = kWinAttributes -0x16 = kComment - -0x17 = kEncodedHeader - -0x18 = kStartPos -0x19 = kDummy - - -7z format headers ------------------ - -SignatureHeader -~~~~~~~~~~~~~~~ - BYTE kSignature[6] = {'7', 'z', 0xBC, 0xAF, 0x27, 0x1C}; - - ArchiveVersion - { - BYTE Major; // now = 0 - BYTE Minor; // now = 4 - }; - - UINT32 StartHeaderCRC; - - StartHeader - { - REAL_UINT64 NextHeaderOffset - REAL_UINT64 NextHeaderSize - UINT32 NextHeaderCRC - } - - -........................... - - -ArchiveProperties -~~~~~~~~~~~~~~~~~ -BYTE NID::kArchiveProperties (0x02) -for (;;) -{ - BYTE PropertyType; - if (aType == 0) - break; - UINT64 PropertySize; - BYTE PropertyData[PropertySize]; -} - - -Digests (NumStreams) -~~~~~~~~~~~~~~~~~~~~~ - BYTE AllAreDefined - if (AllAreDefined == 0) - { - for(NumStreams) - BIT Defined - } - UINT32 CRCs[NumDefined] - - -PackInfo -~~~~~~~~~~~~ - BYTE NID::kPackInfo (0x06) - UINT64 PackPos - UINT64 NumPackStreams - - [] - BYTE NID::kSize (0x09) - UINT64 PackSizes[NumPackStreams] - [] - - [] - BYTE NID::kCRC (0x0A) - PackStreamDigests[NumPackStreams] - [] - - BYTE NID::kEnd - - -Folder -~~~~~~ - UINT64 NumCoders; - for (NumCoders) - { - BYTE - { - 0:3 CodecIdSize - 4: Is Complex Coder - 5: There Are Attributes - 6: Reserved - 7: There are more alternative methods. (Not used anymore, must be 0). - } - BYTE CodecId[CodecIdSize] - if (Is Complex Coder) - { - UINT64 NumInStreams; - UINT64 NumOutStreams; - } - if (There Are Attributes) - { - UINT64 PropertiesSize - BYTE Properties[PropertiesSize] - } - } - - NumBindPairs = NumOutStreamsTotal - 1; - - for (NumBindPairs) - { - UINT64 InIndex; - UINT64 OutIndex; - } - - NumPackedStreams = NumInStreamsTotal - NumBindPairs; - if (NumPackedStreams > 1) - for(NumPackedStreams) - { - UINT64 Index; - }; - - - - -Coders Info -~~~~~~~~~~~ - - BYTE NID::kUnPackInfo (0x07) - - - BYTE NID::kFolder (0x0B) - UINT64 NumFolders - BYTE External - switch(External) - { - case 0: - Folders[NumFolders] - case 1: - UINT64 DataStreamIndex - } - - - BYTE ID::kCodersUnPackSize (0x0C) - for(Folders) - for(Folder.NumOutStreams) - UINT64 UnPackSize; - - - [] - BYTE NID::kCRC (0x0A) - UnPackDigests[NumFolders] - [] - - - - BYTE NID::kEnd - - - -SubStreams Info -~~~~~~~~~~~~~~ - BYTE NID::kSubStreamsInfo; (0x08) - - [] - BYTE NID::kNumUnPackStream; (0x0D) - UINT64 NumUnPackStreamsInFolders[NumFolders]; - [] - - - [] - BYTE NID::kSize (0x09) - UINT64 UnPackSizes[] - [] - - - [] - BYTE NID::kCRC (0x0A) - Digests[Number of streams with unknown CRC] - [] - - - BYTE NID::kEnd - - -Streams Info -~~~~~~~~~~~~ - - [] - PackInfo - [] - - - [] - CodersInfo - [] - - - [] - SubStreamsInfo - [] - - BYTE NID::kEnd - - -FilesInfo -~~~~~~~~~ - BYTE NID::kFilesInfo; (0x05) - UINT64 NumFiles - - for (;;) - { - BYTE PropertyType; - if (aType == 0) - break; - - UINT64 Size; - - switch(PropertyType) - { - kEmptyStream: (0x0E) - for(NumFiles) - BIT IsEmptyStream - - kEmptyFile: (0x0F) - for(EmptyStreams) - BIT IsEmptyFile - - kAnti: (0x10) - for(EmptyStreams) - BIT IsAntiFile - - case kCTime: (0x12) - case kATime: (0x13) - case kMTime: (0x14) - BYTE AllAreDefined - if (AllAreDefined == 0) - { - for(NumFiles) - BIT TimeDefined - } - BYTE External; - if(External != 0) - UINT64 DataIndex - [] - for(Definded Items) - REAL_UINT64 Time - [] - - kNames: (0x11) - BYTE External; - if(External != 0) - UINT64 DataIndex - [] - for(Files) - { - wchar_t Names[NameSize]; - wchar_t 0; - } - [] - - kAttributes: (0x15) - BYTE AllAreDefined - if (AllAreDefined == 0) - { - for(NumFiles) - BIT AttributesAreDefined - } - BYTE External; - if(External != 0) - UINT64 DataIndex - [] - for(Definded Attributes) - UINT32 Attributes - [] - } - } - - -Header -~~~~~~ - BYTE NID::kHeader (0x01) - - [] - ArchiveProperties - [] - - [] - BYTE NID::kAdditionalStreamsInfo; (0x03) - StreamsInfo - [] - - [] - BYTE NID::kMainStreamsInfo; (0x04) - StreamsInfo - [] - - [] - FilesInfo - [] - - BYTE NID::kEnd - - -HeaderInfo -~~~~~~~~~~ - [] - BYTE NID::kEncodedHeader; (0x17) - StreamsInfo for Encoded Header - [] - - ---- -End of document +7z Format description (18.06) +---------------------------- + +This file contains description of 7z archive format. +7z archive can contain files compressed with any method. +See "Methods.txt" for description for defined compressing methods. + + +Format structure Overview +------------------------- + +Some fields can be optional. + +Archive structure +~~~~~~~~~~~~~~~~~ +SignatureHeader +[PackedStreams] +[PackedStreamsForHeaders] +[ + Header + or + { + Packed Header + HeaderInfo + } +] + + + +Header structure +~~~~~~~~~~~~~~~~ +{ + ArchiveProperties + AdditionalStreams + { + PackInfo + { + PackPos + NumPackStreams + Sizes[NumPackStreams] + CRCs[NumPackStreams] + } + CodersInfo + { + NumFolders + Folders[NumFolders] + { + NumCoders + CodersInfo[NumCoders] + { + ID + NumInStreams; + NumOutStreams; + PropertiesSize + Properties[PropertiesSize] + } + NumBindPairs + BindPairsInfo[NumBindPairs] + { + InIndex; + OutIndex; + } + PackedIndices + } + UnPackSize[Folders][Folders.NumOutstreams] + CRCs[NumFolders] + } + SubStreamsInfo + { + NumUnPackStreamsInFolders[NumFolders]; + UnPackSizes[] + CRCs[] + } + } + MainStreamsInfo + { + (Same as in AdditionalStreams) + } + FilesInfo + { + NumFiles + Properties[] + { + ID + Size + Data + } + } +} + +HeaderInfo structure +~~~~~~~~~~~~~~~~~~~~ +{ + (Same as in AdditionalStreams) +} + + + +Notes about Notation and encoding +--------------------------------- + +7z uses little endian encoding. + +7z archive format has optional headers that are marked as +[] +Header +[] + +REAL_UINT64 means real UINT64. + +UINT64 means real UINT64 encoded with the following scheme: + + Size of encoding sequence depends from first byte: + First_Byte Extra_Bytes Value + (binary) + 0xxxxxxx : ( xxxxxxx ) + 10xxxxxx BYTE y[1] : ( xxxxxx << (8 * 1)) + y + 110xxxxx BYTE y[2] : ( xxxxx << (8 * 2)) + y + ... + 1111110x BYTE y[6] : ( x << (8 * 6)) + y + 11111110 BYTE y[7] : y + 11111111 BYTE y[8] : y + + + +Property IDs +------------ + +0x00 = kEnd + +0x01 = kHeader + +0x02 = kArchiveProperties + +0x03 = kAdditionalStreamsInfo +0x04 = kMainStreamsInfo +0x05 = kFilesInfo + +0x06 = kPackInfo +0x07 = kUnPackInfo +0x08 = kSubStreamsInfo + +0x09 = kSize +0x0A = kCRC + +0x0B = kFolder + +0x0C = kCodersUnPackSize +0x0D = kNumUnPackStream + +0x0E = kEmptyStream +0x0F = kEmptyFile +0x10 = kAnti + +0x11 = kName +0x12 = kCTime +0x13 = kATime +0x14 = kMTime +0x15 = kWinAttributes +0x16 = kComment + +0x17 = kEncodedHeader + +0x18 = kStartPos +0x19 = kDummy + + +7z format headers +----------------- + +SignatureHeader +~~~~~~~~~~~~~~~ + BYTE kSignature[6] = {'7', 'z', 0xBC, 0xAF, 0x27, 0x1C}; + + ArchiveVersion + { + BYTE Major; // now = 0 + BYTE Minor; // now = 4 + }; + + UINT32 StartHeaderCRC; + + StartHeader + { + REAL_UINT64 NextHeaderOffset + REAL_UINT64 NextHeaderSize + UINT32 NextHeaderCRC + } + + +........................... + + +ArchiveProperties +~~~~~~~~~~~~~~~~~ +BYTE NID::kArchiveProperties (0x02) +for (;;) +{ + BYTE PropertyType; + if (aType == 0) + break; + UINT64 PropertySize; + BYTE PropertyData[PropertySize]; +} + + +Digests (NumStreams) +~~~~~~~~~~~~~~~~~~~~~ + BYTE AllAreDefined + if (AllAreDefined == 0) + { + for(NumStreams) + BIT Defined + } + UINT32 CRCs[NumDefined] + + +PackInfo +~~~~~~~~~~~~ + BYTE NID::kPackInfo (0x06) + UINT64 PackPos + UINT64 NumPackStreams + + [] + BYTE NID::kSize (0x09) + UINT64 PackSizes[NumPackStreams] + [] + + [] + BYTE NID::kCRC (0x0A) + PackStreamDigests[NumPackStreams] + [] + + BYTE NID::kEnd + + +Folder +~~~~~~ + UINT64 NumCoders; + for (NumCoders) + { + BYTE + { + 0:3 CodecIdSize + 4: Is Complex Coder + 5: There Are Attributes + 6: Reserved + 7: There are more alternative methods. (Not used anymore, must be 0). + } + BYTE CodecId[CodecIdSize] + if (Is Complex Coder) + { + UINT64 NumInStreams; + UINT64 NumOutStreams; + } + if (There Are Attributes) + { + UINT64 PropertiesSize + BYTE Properties[PropertiesSize] + } + } + + NumBindPairs = NumOutStreamsTotal - 1; + + for (NumBindPairs) + { + UINT64 InIndex; + UINT64 OutIndex; + } + + NumPackedStreams = NumInStreamsTotal - NumBindPairs; + if (NumPackedStreams > 1) + for(NumPackedStreams) + { + UINT64 Index; + }; + + + + +Coders Info +~~~~~~~~~~~ + + BYTE NID::kUnPackInfo (0x07) + + + BYTE NID::kFolder (0x0B) + UINT64 NumFolders + BYTE External + switch(External) + { + case 0: + Folders[NumFolders] + case 1: + UINT64 DataStreamIndex + } + + + BYTE ID::kCodersUnPackSize (0x0C) + for(Folders) + for(Folder.NumOutStreams) + UINT64 UnPackSize; + + + [] + BYTE NID::kCRC (0x0A) + UnPackDigests[NumFolders] + [] + + + + BYTE NID::kEnd + + + +SubStreams Info +~~~~~~~~~~~~~~ + BYTE NID::kSubStreamsInfo; (0x08) + + [] + BYTE NID::kNumUnPackStream; (0x0D) + UINT64 NumUnPackStreamsInFolders[NumFolders]; + [] + + + [] + BYTE NID::kSize (0x09) + UINT64 UnPackSizes[] + [] + + + [] + BYTE NID::kCRC (0x0A) + Digests[Number of streams with unknown CRC] + [] + + + BYTE NID::kEnd + + +Streams Info +~~~~~~~~~~~~ + + [] + PackInfo + [] + + + [] + CodersInfo + [] + + + [] + SubStreamsInfo + [] + + BYTE NID::kEnd + + +FilesInfo +~~~~~~~~~ + BYTE NID::kFilesInfo; (0x05) + UINT64 NumFiles + + for (;;) + { + BYTE PropertyType; + if (aType == 0) + break; + + UINT64 Size; + + switch(PropertyType) + { + kEmptyStream: (0x0E) + for(NumFiles) + BIT IsEmptyStream + + kEmptyFile: (0x0F) + for(EmptyStreams) + BIT IsEmptyFile + + kAnti: (0x10) + for(EmptyStreams) + BIT IsAntiFile + + case kCTime: (0x12) + case kATime: (0x13) + case kMTime: (0x14) + BYTE AllAreDefined + if (AllAreDefined == 0) + { + for(NumFiles) + BIT TimeDefined + } + BYTE External; + if(External != 0) + UINT64 DataIndex + [] + for(Definded Items) + REAL_UINT64 Time + [] + + kNames: (0x11) + BYTE External; + if(External != 0) + UINT64 DataIndex + [] + for(Files) + { + wchar_t Names[NameSize]; + wchar_t 0; + } + [] + + kAttributes: (0x15) + BYTE AllAreDefined + if (AllAreDefined == 0) + { + for(NumFiles) + BIT AttributesAreDefined + } + BYTE External; + if(External != 0) + UINT64 DataIndex + [] + for(Definded Attributes) + UINT32 Attributes + [] + } + } + + +Header +~~~~~~ + BYTE NID::kHeader (0x01) + + [] + ArchiveProperties + [] + + [] + BYTE NID::kAdditionalStreamsInfo; (0x03) + StreamsInfo + [] + + [] + BYTE NID::kMainStreamsInfo; (0x04) + StreamsInfo + [] + + [] + FilesInfo + [] + + BYTE NID::kEnd + + +HeaderInfo +~~~~~~~~~~ + [] + BYTE NID::kEncodedHeader; (0x17) + StreamsInfo for Encoded Header + [] + + +--- +End of document diff --git a/deps/LZMA-SDK/DOC/Methods.txt b/deps/LZMA-SDK/DOC/Methods.txt index 6d0641bae..d4a1b1dd4 100644 --- a/deps/LZMA-SDK/DOC/Methods.txt +++ b/deps/LZMA-SDK/DOC/Methods.txt @@ -1,173 +1,173 @@ -7-Zip method IDs for 7z and xz archives ---------------------------------------- - -Version: 18.06 -Date: 2018-06-30 - -Each compression or crypto method in 7z is associated with unique binary value (ID). -The length of ID in bytes is arbitrary but it can not exceed 63 bits (8 bytes). - -xz and 7z formats use same ID map. - -If you want to add some new ID, you have two ways: - 1) Write request for allocating IDs to 7-Zip developers. - 2) Generate 8-bytes ID: - - 3F ZZ ZZ ZZ ZZ ZZ MM MM - - 3F - Prefix for random IDs (1 byte) - ZZ ZZ ZZ ZZ ZZ - Developer ID (5 bytes). Use real random bytes. - - MM MM - Method ID (2 bytes) - - You can notify 7-Zip developers about your Developer ID / Method ID. - - Note: Use new ID, if old codec can not decode data encoded with new version. - - -List of defined IDs -------------------- - -00 - Copy - -03 - Delta -04 - BCJ (x86) -05 - PPC (big-endian) -06 - IA64 -07 - ARM (little-endian) -08 - ARMT (little-endian) -09 - SPARC - -21 - LZMA2 - -02.. - Common - 03 [Swap] - - 2 Swap2 - - 4 Swap4 - -03.. - 7z - 01 - - 01 - LZMA - - 03 - [Branch Codecs] - 01 - [x86 Codecs] - 03 - BCJ - 1B - BCJ2 (4 packed streams) - 02 - - 05 - PPC (big-endian) - 03 - - 01 - Alpha - 04 - - 01 - IA64 - 05 - - 01 - ARM (little-endian) - 06 - - 05 - M68 (big-endian) - 07 - - 01 - ARMT (little-endian) - 08 - - 05 - SPARC - - 04 - - 01 - PPMD - - 7F - - 01 - experimental method. - - -04.. - Misc codecs - - 00 - Reserved - - 01 - [Zip] - 00 - Copy (not used. Use {00} instead) - 01 - Shrink - 06 - Implode - 08 - Deflate - 09 - Deflate64 - 0A - Imploding - 0C - BZip2 (not used. Use {040202} instead) - 0E - LZMA (LZMA-zip) - 5F - xz - 60 - Jpeg - 61 - WavPack - 62 - PPMd (PPMd-zip) - 63 - wzAES - - 02 - - 02 - BZip2 - - 03 - [Rar] - 01 - Rar1 - 02 - Rar2 - 03 - Rar3 - 05 - Rar5 - - 04 - [Arj] - 01 - Arj(1,2,3) - 02 - Arj4 - - 05 - [Z] - - 06 - [Lzh] - - 07 - Reserved for 7z - - 08 - [Cab] - - 09 - [NSIS] - 01 - DeflateNSIS - 02 - BZip2NSIS - - F7 - External codecs (that are not included to 7-Zip) - - 0x xx - reserved - - 10 xx - reserved (LZHAM) - 01 - LZHAM - - 11 xx - reserved (Tino Reichardt) - 01 - ZSTD - 02 - BROTLI - 04 - LZ4 - 05 - LZ5 - 06 - LIZARD - - 12 xx - reserverd (Denis Anisimov) - - 01 - WavPack2 - FE - eSplitter - FF - RawSplitter - - -06.. - Crypto - - F0 - Ciphers without hashing algo - - 01 - [AES] - 0x - AES-128 - 4x - AES-192 - 8x - AES-256 - Cx - AES - - x0 - ECB - x1 - CBC - x2 - CFB - x3 - OFB - x4 - CTR - - F1 - Combine Ciphers - - 01 - [Zip] - 01 - ZipCrypto (Main Zip crypto algo) - - 03 - [RAR] - 02 - - 03 - Rar29AES (AES-128 + modified SHA-1) - - 07 - [7z] - 01 - 7zAES (AES-256 + SHA-256) - - ---- -End of document +7-Zip method IDs for 7z and xz archives +--------------------------------------- + +Version: 18.06 +Date: 2018-06-30 + +Each compression or crypto method in 7z is associated with unique binary value (ID). +The length of ID in bytes is arbitrary but it can not exceed 63 bits (8 bytes). + +xz and 7z formats use same ID map. + +If you want to add some new ID, you have two ways: + 1) Write request for allocating IDs to 7-Zip developers. + 2) Generate 8-bytes ID: + + 3F ZZ ZZ ZZ ZZ ZZ MM MM + + 3F - Prefix for random IDs (1 byte) + ZZ ZZ ZZ ZZ ZZ - Developer ID (5 bytes). Use real random bytes. + + MM MM - Method ID (2 bytes) + + You can notify 7-Zip developers about your Developer ID / Method ID. + + Note: Use new ID, if old codec can not decode data encoded with new version. + + +List of defined IDs +------------------- + +00 - Copy + +03 - Delta +04 - BCJ (x86) +05 - PPC (big-endian) +06 - IA64 +07 - ARM (little-endian) +08 - ARMT (little-endian) +09 - SPARC + +21 - LZMA2 + +02.. - Common + 03 [Swap] + - 2 Swap2 + - 4 Swap4 + +03.. - 7z + 01 - + 01 - LZMA + + 03 - [Branch Codecs] + 01 - [x86 Codecs] + 03 - BCJ + 1B - BCJ2 (4 packed streams) + 02 - + 05 - PPC (big-endian) + 03 - + 01 - Alpha + 04 - + 01 - IA64 + 05 - + 01 - ARM (little-endian) + 06 - + 05 - M68 (big-endian) + 07 - + 01 - ARMT (little-endian) + 08 - + 05 - SPARC + + 04 - + 01 - PPMD + + 7F - + 01 - experimental method. + + +04.. - Misc codecs + + 00 - Reserved + + 01 - [Zip] + 00 - Copy (not used. Use {00} instead) + 01 - Shrink + 06 - Implode + 08 - Deflate + 09 - Deflate64 + 0A - Imploding + 0C - BZip2 (not used. Use {040202} instead) + 0E - LZMA (LZMA-zip) + 5F - xz + 60 - Jpeg + 61 - WavPack + 62 - PPMd (PPMd-zip) + 63 - wzAES + + 02 - + 02 - BZip2 + + 03 - [Rar] + 01 - Rar1 + 02 - Rar2 + 03 - Rar3 + 05 - Rar5 + + 04 - [Arj] + 01 - Arj(1,2,3) + 02 - Arj4 + + 05 - [Z] + + 06 - [Lzh] + + 07 - Reserved for 7z + + 08 - [Cab] + + 09 - [NSIS] + 01 - DeflateNSIS + 02 - BZip2NSIS + + F7 - External codecs (that are not included to 7-Zip) + + 0x xx - reserved + + 10 xx - reserved (LZHAM) + 01 - LZHAM + + 11 xx - reserved (Tino Reichardt) + 01 - ZSTD + 02 - BROTLI + 04 - LZ4 + 05 - LZ5 + 06 - LIZARD + + 12 xx - reserverd (Denis Anisimov) + + 01 - WavPack2 + FE - eSplitter + FF - RawSplitter + + +06.. - Crypto + + F0 - Ciphers without hashing algo + + 01 - [AES] + 0x - AES-128 + 4x - AES-192 + 8x - AES-256 + Cx - AES + + x0 - ECB + x1 - CBC + x2 - CFB + x3 - OFB + x4 - CTR + + F1 - Combine Ciphers + + 01 - [Zip] + 01 - ZipCrypto (Main Zip crypto algo) + + 03 - [RAR] + 02 - + 03 - Rar29AES (AES-128 + modified SHA-1) + + 07 - [7z] + 01 - 7zAES (AES-256 + SHA-256) + + +--- +End of document diff --git a/deps/LZMA-SDK/DOC/installer.txt b/deps/LZMA-SDK/DOC/installer.txt index 70ad7dc6a..b99d21d24 100644 --- a/deps/LZMA-SDK/DOC/installer.txt +++ b/deps/LZMA-SDK/DOC/installer.txt @@ -1,166 +1,166 @@ -7-Zip for installers 9.38 -------------------------- - -7-Zip is a file archiver for Windows NT/2000/2003/2008/XP/Vista/7/8/10. - -7-Zip for installers is part of LZMA SDK. -LZMA SDK is written and placed in the public domain by Igor Pavlov. - -It's allowed to join 7-Zip SFX module with another software. -It's allowed to change resources of 7-Zip's SFX modules. - - -HOW to use ------------ - -7zr.exe is reduced version of 7za.exe of 7-Zip. -7zr.exe supports only format with these codecs: LZMA, LZMA2, BCJ, BCJ2, ARM, Copy. - -Example of compressing command for installation packages: - -7zr a archive.7z files - -7zSD.sfx is SFX module for installers. 7zSD.sfx uses msvcrt.dll. - -SFX modules for installers allow to create installation program. -Such module extracts archive to temp folder and then runs specified program and removes -temp files after program finishing. Self-extract archive for installers must be created -as joining 3 files: SFX_Module, Installer_Config, 7z_Archive. -Installer_Config is optional file. You can use the following command to create installer -self-extract archive: - -copy /b 7zSD.sfx + config.txt + archive.7z archive.exe - -The smallest installation package size can be achieved, if installation files was -uncompressed before including to 7z archive. - --y switch for installer module (at runtime) specifies quiet mode for extracting. - -Installer Config file format -~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -Config file contains commands for Installer. File begins from string -;!@Install@!UTF-8! and ends with ;!@InstallEnd@!. File must be written -in UTF-8 encoding. File contains string pairs: - -ID_String="Value" - -ID_String Description - -Title Title for messages -BeginPrompt Begin Prompt message -Progress Value can be "yes" or "no". Default value is "yes". -RunProgram Command for executing. Default value is "setup.exe". - Substring %%T will be replaced with path to temporary - folder, where files were extracted -Directory Directory prefix for "RunProgram". Default value is ".\\" -ExecuteFile Name of file for executing -ExecuteParameters Parameters for "ExecuteFile" - - -You can omit any string pair. - -There are two ways to run program: RunProgram and ExecuteFile. -Use RunProgram, if you want to run some program from .7z archive. -Use ExecuteFile, if you want to open some document from .7z archive or -if you want to execute some command from Windows. - -If you use RunProgram and if you specify empty directory prefix: Directory="", -the system searches for the executable file in the following sequence: - -1. The directory from which the application (installer) loaded. -2. The temporary folder, where files were extracted. -3. The Windows system directory. - - -Config file Examples -~~~~~~~~~~~~~~~~~~~~ - -;!@Install@!UTF-8! -Title="7-Zip 4.00" -BeginPrompt="Do you want to install the 7-Zip 4.00?" -RunProgram="setup.exe" -;!@InstallEnd@! - - - -;!@Install@!UTF-8! -Title="7-Zip 4.00" -BeginPrompt="Do you want to install the 7-Zip 4.00?" -ExecuteFile="7zip.msi" -;!@InstallEnd@! - - - -;!@Install@!UTF-8! -Title="7-Zip 4.01 Update" -BeginPrompt="Do you want to install the 7-Zip 4.01 Update?" -ExecuteFile="msiexec.exe" -ExecuteParameters="/i 7zip.msi REINSTALL=ALL REINSTALLMODE=vomus" -;!@InstallEnd@! - - - -Small SFX modules for installers --------------------------------- - -7zS2.sfx - small SFX module (GUI version) -7zS2con.sfx - small SFX module (Console version) - -Small SFX modules support this codecs: LZMA, LZMA2, BCJ, BCJ2, ARM, COPY - -Small SFX module is similar to common SFX module for installers. -The difference (what's new in small version): - - Smaller size (30 KB vs 100 KB) - - C source code instead of Ñ++ - - No installer Configuration file - - No extracting progress window - - It decompresses solid 7z blocks (it can be whole 7z archive) to RAM. - So user that calls SFX installer must have free RAM of size of largest - solid 7z block (size of 7z archive at simplest case). - -How to use ----------- - -copy /b 7zS2.sfx + archive.7z sfx.exe - -When you run installer sfx module (sfx.exe) -1) It creates "7zNNNNNNNN" temp folder in system temp folder. -2) It extracts .7z archive to that folder -3) It executes one file from "7zNNNNNNNN" temp folder. -4) It removes "7zNNNNNNNN" temp folder - -You can send parameters to installer, and installer will transfer them to extracted .exe file. - -Small SFX uses 3 levels of priorities to select file to execute: - - 1) Files in root folder have higher priority than files in subfolders. - 2) File extension priorities (from high to low priority order): - bat, cmd, exe, inf, msi, cab (under Windows CE), html, htm - 3) File name priorities (from high to low priority order): - setup, install, run, start - -Windows CE (ARM) version of 7zS2.sfx is included to 7-Zip for Windows Mobile package. - - -Examples --------- - -1) To create compressed console 7-Zip: - -7zr a c.7z 7z.exe 7z.dll -mx -copy /b 7zS2con.sfx + c.7z 7zCompr.exe -7zCompr.exe b -md22 - - -2) To create compressed GUI 7-Zip: - -7zr a g.7z 7zg.exe 7z.dll -mx -copy /b 7zS2.sfx + g.7z 7zgCompr.exe -7zgCompr.exe b -md22 - - -3) To open some file: - -7zr a h.7z readme.txt -mx -copy /b 7zS2.sfx + h.7z 7zTxt.exe -7zTxt.exe +7-Zip for installers 9.38 +------------------------- + +7-Zip is a file archiver for Windows NT/2000/2003/2008/XP/Vista/7/8/10. + +7-Zip for installers is part of LZMA SDK. +LZMA SDK is written and placed in the public domain by Igor Pavlov. + +It's allowed to join 7-Zip SFX module with another software. +It's allowed to change resources of 7-Zip's SFX modules. + + +HOW to use +----------- + +7zr.exe is reduced version of 7za.exe of 7-Zip. +7zr.exe supports only format with these codecs: LZMA, LZMA2, BCJ, BCJ2, ARM, Copy. + +Example of compressing command for installation packages: + +7zr a archive.7z files + +7zSD.sfx is SFX module for installers. 7zSD.sfx uses msvcrt.dll. + +SFX modules for installers allow to create installation program. +Such module extracts archive to temp folder and then runs specified program and removes +temp files after program finishing. Self-extract archive for installers must be created +as joining 3 files: SFX_Module, Installer_Config, 7z_Archive. +Installer_Config is optional file. You can use the following command to create installer +self-extract archive: + +copy /b 7zSD.sfx + config.txt + archive.7z archive.exe + +The smallest installation package size can be achieved, if installation files was +uncompressed before including to 7z archive. + +-y switch for installer module (at runtime) specifies quiet mode for extracting. + +Installer Config file format +~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Config file contains commands for Installer. File begins from string +;!@Install@!UTF-8! and ends with ;!@InstallEnd@!. File must be written +in UTF-8 encoding. File contains string pairs: + +ID_String="Value" + +ID_String Description + +Title Title for messages +BeginPrompt Begin Prompt message +Progress Value can be "yes" or "no". Default value is "yes". +RunProgram Command for executing. Default value is "setup.exe". + Substring %%T will be replaced with path to temporary + folder, where files were extracted +Directory Directory prefix for "RunProgram". Default value is ".\\" +ExecuteFile Name of file for executing +ExecuteParameters Parameters for "ExecuteFile" + + +You can omit any string pair. + +There are two ways to run program: RunProgram and ExecuteFile. +Use RunProgram, if you want to run some program from .7z archive. +Use ExecuteFile, if you want to open some document from .7z archive or +if you want to execute some command from Windows. + +If you use RunProgram and if you specify empty directory prefix: Directory="", +the system searches for the executable file in the following sequence: + +1. The directory from which the application (installer) loaded. +2. The temporary folder, where files were extracted. +3. The Windows system directory. + + +Config file Examples +~~~~~~~~~~~~~~~~~~~~ + +;!@Install@!UTF-8! +Title="7-Zip 4.00" +BeginPrompt="Do you want to install the 7-Zip 4.00?" +RunProgram="setup.exe" +;!@InstallEnd@! + + + +;!@Install@!UTF-8! +Title="7-Zip 4.00" +BeginPrompt="Do you want to install the 7-Zip 4.00?" +ExecuteFile="7zip.msi" +;!@InstallEnd@! + + + +;!@Install@!UTF-8! +Title="7-Zip 4.01 Update" +BeginPrompt="Do you want to install the 7-Zip 4.01 Update?" +ExecuteFile="msiexec.exe" +ExecuteParameters="/i 7zip.msi REINSTALL=ALL REINSTALLMODE=vomus" +;!@InstallEnd@! + + + +Small SFX modules for installers +-------------------------------- + +7zS2.sfx - small SFX module (GUI version) +7zS2con.sfx - small SFX module (Console version) + +Small SFX modules support this codecs: LZMA, LZMA2, BCJ, BCJ2, ARM, COPY + +Small SFX module is similar to common SFX module for installers. +The difference (what's new in small version): + - Smaller size (30 KB vs 100 KB) + - C source code instead of Ñ++ + - No installer Configuration file + - No extracting progress window + - It decompresses solid 7z blocks (it can be whole 7z archive) to RAM. + So user that calls SFX installer must have free RAM of size of largest + solid 7z block (size of 7z archive at simplest case). + +How to use +---------- + +copy /b 7zS2.sfx + archive.7z sfx.exe + +When you run installer sfx module (sfx.exe) +1) It creates "7zNNNNNNNN" temp folder in system temp folder. +2) It extracts .7z archive to that folder +3) It executes one file from "7zNNNNNNNN" temp folder. +4) It removes "7zNNNNNNNN" temp folder + +You can send parameters to installer, and installer will transfer them to extracted .exe file. + +Small SFX uses 3 levels of priorities to select file to execute: + + 1) Files in root folder have higher priority than files in subfolders. + 2) File extension priorities (from high to low priority order): + bat, cmd, exe, inf, msi, cab (under Windows CE), html, htm + 3) File name priorities (from high to low priority order): + setup, install, run, start + +Windows CE (ARM) version of 7zS2.sfx is included to 7-Zip for Windows Mobile package. + + +Examples +-------- + +1) To create compressed console 7-Zip: + +7zr a c.7z 7z.exe 7z.dll -mx +copy /b 7zS2con.sfx + c.7z 7zCompr.exe +7zCompr.exe b -md22 + + +2) To create compressed GUI 7-Zip: + +7zr a g.7z 7zg.exe 7z.dll -mx +copy /b 7zS2.sfx + g.7z 7zgCompr.exe +7zgCompr.exe b -md22 + + +3) To open some file: + +7zr a h.7z readme.txt -mx +copy /b 7zS2.sfx + h.7z 7zTxt.exe +7zTxt.exe diff --git a/deps/LZMA-SDK/DOC/lzma-history.txt b/deps/LZMA-SDK/DOC/lzma-history.txt index 9c5a2416c..0963c7bf7 100644 --- a/deps/LZMA-SDK/DOC/lzma-history.txt +++ b/deps/LZMA-SDK/DOC/lzma-history.txt @@ -1,520 +1,520 @@ -HISTORY of the LZMA SDK ------------------------ - -21.07 2021-12-26 -------------------------- -- New switches: -spm and -im!{file_path} to exclude directories from processing - for specified paths that don't contain path separator character at the end of path. -- The sorting order of files in archives was slightly changed to be more consistent - for cases where the name of some directory is the same as the prefix part of the name - of another directory or file. - - -21.06 2021-11-24 -------------------------- -- Bug in LZMA encoder in file LzmaEnc.c was fixed: - LzmaEnc_MemEncode(), LzmaEncode() and LzmaCompress() could work incorrectly, - if size value for output buffer is smaller than size required for all compressed data. - LzmaEnc_Encode() could work incorrectly, - if callback ISeqOutStream::Write() doesn't write all compressed data. - NCompress::NLzma::CEncoder::Code() could work incorrectly, - if callback ISequentialOutStream::Write() returns error code. -- Bug in versions 21.00-21.05 was fixed: - 7-Zip didn't set attributes of directories during archive extracting. - - -21.04 beta 2021-11-02 -------------------------- -- 7-Zip now reduces the number of working CPU threads for compression, - if RAM size is not enough for compression with big LZMA2 dictionary. -- 7-Zip now can create and check "file.sha256" text files that contain the list - of file names and SHA-256 checksums in format compatible with sha256sum program. - - -21.03 beta 2021-07-20 -------------------------- -- The maximum dictionary size for LZMA/LZMA2 compressing was increased to 4 GB (3840 MiB). -- Minor speed optimizations in LZMA/LZMA2 compressing. - - -21.02 alpha 2021-05-06 -------------------------- -- The command line version of 7-Zip for macOS was released. -- The speed for LZMA and LZMA2 decompression in arm64 versions for macOS and Linux - was increased by 20%-60%. - - -21.01 alpha 2021-03-09 -------------------------- -- The command line version of 7-Zip for Linux was released. -- The improvements for speed of ARM64 version using hardware CPU instructions - for AES, CRC-32, SHA-1 and SHA-256. -- Some bugs were fixed. - - -20.02 alpha 2020-08-08 -------------------------- -- The default number of LZMA2 chunks per solid block in 7z archive was increased to 64. - It allows to increase the compression speed for big 7z archives, if there is a big number - of CPU cores and threads. -- The speed of PPMd compressing/decompressing was increased for 7z archives. -- The new -ssp switch. If the switch -ssp is specified, 7-Zip doesn't allow the system - to modify "Last Access Time" property of source files for archiving and hashing operations. -- Some bugs were fixed. - - -20.00 alpha 2020-02-06 -------------------------- -- 7-Zip now supports new optional match finders for LZMA/LZMA2 compression: bt5 and hc5, - that can work faster than bt4 and hc4 match finders for the data with big redundancy. -- The compression ratio was improved for Fast and Fastest compression levels with the - following default settings: - - Fastest level (-mx1) : hc5 match finder with 256 KB dictionary. - - Fast level (-mx3) : hc5 match finder with 4 MB dictionary. -- Minor speed optimizations in multithreaded LZMA/LZMA2 compression for Normal/Maximum/Ultra - compression levels. - - -19.00 2019-02-21 -------------------------- -- Encryption strength for 7z archives was increased: - the size of random initialization vector was increased from 64-bit to 128-bit, - and the pseudo-random number generator was improved. -- The bug in 7zIn.c code was fixed. - - -18.06 2018-12-30 -------------------------- -- The speed for LZMA/LZMA2 compressing was increased by 3-10%, - and there are minor changes in compression ratio. -- Some bugs were fixed. -- The bug in 7-Zip 18.02-18.05 was fixed: - There was memory leak in multithreading xz decoder - XzDecMt_Decode(), - if xz stream contains only one block. -- The changes for MSVS compiler makefiles: - - the makefiles now use "PLATFORM" macroname with values (x64, x86, arm64) - instead of "CPU" macroname with values (AMD64, ARM64). - - the makefiles by default now use static version of the run-time library. - - -18.05 2018-04-30 -------------------------- -- The speed for LZMA/LZMA2 compressing was increased - by 8% for fastest/fast compression levels and - by 3% for normal/maximum compression levels. -- Previous versions of 7-Zip could work incorrectly in "Large memory pages" mode in - Windows 10 because of some BUG with "Large Pages" in Windows 10. - Now 7-Zip doesn't use "Large Pages" on Windows 10 up to revision 1709 (16299). -- The BUG was fixed in Lzma2Enc.c - Lzma2Enc_Encode2() function worked incorretly, - if (inStream == NULL) and the number of block threads is more than 1. - - -18.03 beta 2018-03-04 -------------------------- -- Asm\x86\LzmaDecOpt.asm: new optimized LZMA decoder written in asm - for x64 with about 30% higher speed than main version of LZMA decoder written in C. -- The speed for single-thread LZMA/LZMA2 decoder written in C was increased by 3%. -- 7-Zip now can use multi-threading for 7z/LZMA2 decoding, - if there are multiple independent data chunks in LZMA2 stream. -- 7-Zip now can use multi-threading for xz decoding, - if there are multiple blocks in xz stream. - - -18.01 2019-01-28 -------------------------- -- The BUG in 17.01 - 18.00 beta was fixed: - XzDec.c : random block unpacking and XzUnpacker_IsBlockFinished() - didn't work correctly for xz archives without checksum (CRC). - - -18.00 beta 2019-01-10 -------------------------- -- The BUG in xz encoder was fixed: - There was memory leak of 16 KB for each file compressed with - xz compression method, if additional filter was used. - - -17.01 beta 2017-08-28 -------------------------- -- Minor speed optimization for LZMA2 (xz and 7z) multi-threading compression. - 7-Zip now uses additional memory buffers for multi-block LZMA2 compression. - CPU utilization was slightly improved. -- 7-zip now creates multi-block xz archives by default. Block size can be - specified with -ms[Size]{m|g} switch. -- xz decoder now can unpack random block from multi-block xz archives. -- 7-Zip command line: @listfile now doesn't work after -- switch. - Use -i@listfile before -- switch instead. -- The BUGs were fixed: - 7-Zip 17.00 beta crashed for commands that write anti-item to 7z archive. - - -17.00 beta 2017-04-29 -------------------------- -- NewHandler.h / NewHandler.cpp: - now it redefines operator new() only for old MSVC compilers (_MSC_VER < 1900). -- C/7zTypes.h : the names of variables in interface structures were changed (vt). -- Some bugs were fixed. 7-Zip could crash in some cases. -- Some internal changes in code. - - -16.04 2016-10-04 -------------------------- -- The bug was fixed in DllSecur.c. - - -16.03 2016-09-28 -------------------------- -- SFX modules now use some protection against DLL preloading attack. -- Some bugs in 7z code were fixed. - - -16.02 2016-05-21 -------------------------- -- The BUG in 16.00 - 16.01 was fixed: - Split Handler (SplitHandler.cpp) returned incorrect - total size value (kpidSize) for split archives. - - -16.01 2016-05-19 -------------------------- -- Some internal changes to reduce the number of compiler warnings. - - -16.00 2016-05-10 -------------------------- -- Some bugs were fixed. - - -15.12 2015-11-19 -------------------------- -- The BUG in C version of 7z decoder was fixed: - 7zDec.c : SzDecodeLzma2() - 7z decoder could mistakenly report about decoding error for some 7z archives - that use LZMA2 compression method. - The probability to get that mistaken decoding error report was about - one error per 16384 solid blocks for solid blocks larger than 16 KB (compressed size). -- The BUG (in 9.26-15.11) in C version of 7z decoder was fixed: - 7zArcIn.c : SzReadHeader2() - 7z decoder worked incorrectly for 7z archives that contain - empty solid blocks, that can be placed to 7z archive, if some file is - unavailable for reading during archive creation. - - -15.09 beta 2015-10-16 -------------------------- -- The BUG in LZMA / LZMA2 encoding code was fixed. - The BUG in LzFind.c::MatchFinder_ReadBlock() function. - If input data size is larger than (4 GiB - dictionary_size), - the following code worked incorrectly: - - LZMA : LzmaEnc_MemEncode(), LzmaEncode() : LZMA encoding functions - for compressing from memory to memory. - That BUG is not related to LZMA encoder version that works via streams. - - LZMA2 : multi-threaded version of LZMA2 encoder worked incorrectly, if - default value of chunk size (CLzma2EncProps::blockSize) is changed - to value larger than (4 GiB - dictionary_size). - - -9.38 beta 2015-01-03 -------------------------- -- The BUG in 9.31-9.37 was fixed: - IArchiveGetRawProps interface was disabled for 7z archives. -- The BUG in 9.26-9.36 was fixed: - Some code in CPP\7zip\Archive\7z\ worked correctly only under Windows. - - -9.36 beta 2014-12-26 -------------------------- -- The BUG in command line version was fixed: - 7-Zip created temporary archive in current folder during update archive - operation, if -w{Path} switch was not specified. - The fixed 7-Zip creates temporary archive in folder that contains updated archive. -- The BUG in 9.33-9.35 was fixed: - 7-Zip silently ignored file reading errors during 7z or gz archive creation, - and the created archive contained only part of file that was read before error. - The fixed 7-Zip stops archive creation and it reports about error. - - -9.35 beta 2014-12-07 -------------------------- -- 7zr.exe now support AES encryption. -- SFX mudules were added to LZMA SDK -- Some bugs were fixed. - - -9.21 beta 2011-04-11 -------------------------- -- New class FString for file names at file systems. -- Speed optimization in CRC code for big-endian CPUs. -- The BUG in Lzma2Dec.c was fixed: - Lzma2Decode function didn't work. - - -9.18 beta 2010-11-02 -------------------------- -- New small SFX module for installers (SfxSetup). - - -9.12 beta 2010-03-24 -------------------------- -- The BUG in LZMA SDK 9.* was fixed: LZMA2 codec didn't work, - if more than 10 threads were used (or more than 20 threads in some modes). - - -9.11 beta 2010-03-15 -------------------------- -- PPMd compression method support - - -9.09 2009-12-12 -------------------------- -- The bug was fixed: - Utf16_To_Utf8 funstions in UTFConvert.cpp and 7zMain.c - incorrectly converted surrogate characters (the code >= 0x10000) to UTF-8. -- Some bugs were fixed - - -9.06 2009-08-17 -------------------------- -- Some changes in ANSI-C 7z Decoder interfaces. - - -9.04 2009-05-30 -------------------------- -- LZMA2 compression method support -- xz format support - - -4.65 2009-02-03 -------------------------- -- Some minor fixes - - -4.63 2008-12-31 -------------------------- -- Some minor fixes - - -4.61 beta 2008-11-23 -------------------------- -- The bug in ANSI-C LZMA Decoder was fixed: - If encoded stream was corrupted, decoder could access memory - outside of allocated range. -- Some changes in ANSI-C 7z Decoder interfaces. -- LZMA SDK is placed in the public domain. - - -4.60 beta 2008-08-19 -------------------------- -- Some minor fixes. - - -4.59 beta 2008-08-13 -------------------------- -- The bug was fixed: - LZMA Encoder in fast compression mode could access memory outside of - allocated range in some rare cases. - - -4.58 beta 2008-05-05 -------------------------- -- ANSI-C LZMA Decoder was rewritten for speed optimizations. -- ANSI-C LZMA Encoder was included to LZMA SDK. -- C++ LZMA code now is just wrapper over ANSI-C code. - - -4.57 2007-12-12 -------------------------- -- Speed optimizations in Ñ++ LZMA Decoder. -- Small changes for more compatibility with some C/C++ compilers. - - -4.49 beta 2007-07-05 -------------------------- -- .7z ANSI-C Decoder: - - now it supports BCJ and BCJ2 filters - - now it supports files larger than 4 GB. - - now it supports "Last Write Time" field for files. -- C++ code for .7z archives compressing/decompressing from 7-zip - was included to LZMA SDK. - - -4.43 2006-06-04 -------------------------- -- Small changes for more compatibility with some C/C++ compilers. - - -4.42 2006-05-15 -------------------------- -- Small changes in .h files in ANSI-C version. - - -4.39 beta 2006-04-14 -------------------------- -- The bug in versions 4.33b:4.38b was fixed: - C++ version of LZMA encoder could not correctly compress - files larger than 2 GB with HC4 match finder (-mfhc4). - - -4.37 beta 2005-04-06 -------------------------- -- Fixes in C++ code: code could no be compiled if _NO_EXCEPTIONS was defined. - - -4.35 beta 2005-03-02 -------------------------- -- The bug was fixed in C++ version of LZMA Decoder: - If encoded stream was corrupted, decoder could access memory - outside of allocated range. - - -4.34 beta 2006-02-27 -------------------------- -- Compressing speed and memory requirements for compressing were increased -- LZMA now can use only these match finders: HC4, BT2, BT3, BT4 - - -4.32 2005-12-09 -------------------------- -- Java version of LZMA SDK was included - - -4.30 2005-11-20 -------------------------- -- Compression ratio was improved in -a2 mode -- Speed optimizations for compressing in -a2 mode -- -fb switch now supports values up to 273 -- The bug in 7z_C (7zIn.c) was fixed: - It used Alloc/Free functions from different memory pools. - So if program used two memory pools, it worked incorrectly. -- 7z_C: .7z format supporting was improved -- LZMA# SDK (C#.NET version) was included - - -4.27 (Updated) 2005-09-21 -------------------------- -- Some GUIDs/interfaces in C++ were changed. - IStream.h: - ISequentialInStream::Read now works as old ReadPart - ISequentialOutStream::Write now works as old WritePart - - -4.27 2005-08-07 -------------------------- -- The bug in LzmaDecodeSize.c was fixed: - if _LZMA_IN_CB and _LZMA_OUT_READ were defined, - decompressing worked incorrectly. - - -4.26 2005-08-05 -------------------------- -- Fixes in 7z_C code and LzmaTest.c: - previous versions could work incorrectly, - if malloc(0) returns 0 - - -4.23 2005-06-29 -------------------------- -- Small fixes in C++ code - - -4.22 2005-06-10 -------------------------- -- Small fixes - - -4.21 2005-06-08 -------------------------- -- Interfaces for ANSI-C LZMA Decoder (LzmaDecode.c) were changed -- New additional version of ANSI-C LZMA Decoder with zlib-like interface: - - LzmaStateDecode.h - - LzmaStateDecode.c - - LzmaStateTest.c -- ANSI-C LZMA Decoder now can decompress files larger than 4 GB - - -4.17 2005-04-18 -------------------------- -- New example for RAM->RAM compressing/decompressing: - LZMA + BCJ (filter for x86 code): - - LzmaRam.h - - LzmaRam.cpp - - LzmaRamDecode.h - - LzmaRamDecode.c - - -f86 switch for lzma.exe - - -4.16 2005-03-29 -------------------------- -- The bug was fixed in LzmaDecode.c (ANSI-C LZMA Decoder): - If _LZMA_OUT_READ was defined, and if encoded stream was corrupted, - decoder could access memory outside of allocated range. -- Speed optimization of ANSI-C LZMA Decoder (now it's about 20% faster). - Old version of LZMA Decoder now is in file LzmaDecodeSize.c. - LzmaDecodeSize.c can provide slightly smaller code than LzmaDecode.c -- Small speed optimization in LZMA C++ code -- filter for SPARC's code was added -- Simplified version of .7z ANSI-C Decoder was included - - -4.06 2004-09-05 -------------------------- -- The bug in v4.05 was fixed: - LZMA-Encoder didn't release output stream in some cases. - - -4.05 2004-08-25 -------------------------- -- Source code of filters for x86, IA-64, ARM, ARM-Thumb - and PowerPC code was included to SDK -- Some internal minor changes - - -4.04 2004-07-28 -------------------------- -- More compatibility with some C++ compilers - - -4.03 2004-06-18 -------------------------- -- "Benchmark" command was added. It measures compressing - and decompressing speed and shows rating values. - Also it checks hardware errors. - - -4.02 2004-06-10 -------------------------- -- C++ LZMA Encoder/Decoder code now is more portable - and it can be compiled by GCC on Linux. - - -4.01 2004-02-15 -------------------------- -- Some detection of data corruption was enabled. - LzmaDecode.c / RangeDecoderReadByte - ..... - { - rd->ExtraBytes = 1; - return 0xFF; - } - - -4.00 2004-02-13 -------------------------- -- Original version of LZMA SDK - - - -HISTORY of the LZMA -------------------- - 2001-2008: Improvements to LZMA compressing/decompressing code, - keeping compatibility with original LZMA format - 1996-2001: Development of LZMA compression format - - Some milestones: - - 2001-08-30: LZMA compression was added to 7-Zip - 1999-01-02: First version of 7-Zip was released - - -End of document +HISTORY of the LZMA SDK +----------------------- + +21.07 2021-12-26 +------------------------- +- New switches: -spm and -im!{file_path} to exclude directories from processing + for specified paths that don't contain path separator character at the end of path. +- The sorting order of files in archives was slightly changed to be more consistent + for cases where the name of some directory is the same as the prefix part of the name + of another directory or file. + + +21.06 2021-11-24 +------------------------- +- Bug in LZMA encoder in file LzmaEnc.c was fixed: + LzmaEnc_MemEncode(), LzmaEncode() and LzmaCompress() could work incorrectly, + if size value for output buffer is smaller than size required for all compressed data. + LzmaEnc_Encode() could work incorrectly, + if callback ISeqOutStream::Write() doesn't write all compressed data. + NCompress::NLzma::CEncoder::Code() could work incorrectly, + if callback ISequentialOutStream::Write() returns error code. +- Bug in versions 21.00-21.05 was fixed: + 7-Zip didn't set attributes of directories during archive extracting. + + +21.04 beta 2021-11-02 +------------------------- +- 7-Zip now reduces the number of working CPU threads for compression, + if RAM size is not enough for compression with big LZMA2 dictionary. +- 7-Zip now can create and check "file.sha256" text files that contain the list + of file names and SHA-256 checksums in format compatible with sha256sum program. + + +21.03 beta 2021-07-20 +------------------------- +- The maximum dictionary size for LZMA/LZMA2 compressing was increased to 4 GB (3840 MiB). +- Minor speed optimizations in LZMA/LZMA2 compressing. + + +21.02 alpha 2021-05-06 +------------------------- +- The command line version of 7-Zip for macOS was released. +- The speed for LZMA and LZMA2 decompression in arm64 versions for macOS and Linux + was increased by 20%-60%. + + +21.01 alpha 2021-03-09 +------------------------- +- The command line version of 7-Zip for Linux was released. +- The improvements for speed of ARM64 version using hardware CPU instructions + for AES, CRC-32, SHA-1 and SHA-256. +- Some bugs were fixed. + + +20.02 alpha 2020-08-08 +------------------------- +- The default number of LZMA2 chunks per solid block in 7z archive was increased to 64. + It allows to increase the compression speed for big 7z archives, if there is a big number + of CPU cores and threads. +- The speed of PPMd compressing/decompressing was increased for 7z archives. +- The new -ssp switch. If the switch -ssp is specified, 7-Zip doesn't allow the system + to modify "Last Access Time" property of source files for archiving and hashing operations. +- Some bugs were fixed. + + +20.00 alpha 2020-02-06 +------------------------- +- 7-Zip now supports new optional match finders for LZMA/LZMA2 compression: bt5 and hc5, + that can work faster than bt4 and hc4 match finders for the data with big redundancy. +- The compression ratio was improved for Fast and Fastest compression levels with the + following default settings: + - Fastest level (-mx1) : hc5 match finder with 256 KB dictionary. + - Fast level (-mx3) : hc5 match finder with 4 MB dictionary. +- Minor speed optimizations in multithreaded LZMA/LZMA2 compression for Normal/Maximum/Ultra + compression levels. + + +19.00 2019-02-21 +------------------------- +- Encryption strength for 7z archives was increased: + the size of random initialization vector was increased from 64-bit to 128-bit, + and the pseudo-random number generator was improved. +- The bug in 7zIn.c code was fixed. + + +18.06 2018-12-30 +------------------------- +- The speed for LZMA/LZMA2 compressing was increased by 3-10%, + and there are minor changes in compression ratio. +- Some bugs were fixed. +- The bug in 7-Zip 18.02-18.05 was fixed: + There was memory leak in multithreading xz decoder - XzDecMt_Decode(), + if xz stream contains only one block. +- The changes for MSVS compiler makefiles: + - the makefiles now use "PLATFORM" macroname with values (x64, x86, arm64) + instead of "CPU" macroname with values (AMD64, ARM64). + - the makefiles by default now use static version of the run-time library. + + +18.05 2018-04-30 +------------------------- +- The speed for LZMA/LZMA2 compressing was increased + by 8% for fastest/fast compression levels and + by 3% for normal/maximum compression levels. +- Previous versions of 7-Zip could work incorrectly in "Large memory pages" mode in + Windows 10 because of some BUG with "Large Pages" in Windows 10. + Now 7-Zip doesn't use "Large Pages" on Windows 10 up to revision 1709 (16299). +- The BUG was fixed in Lzma2Enc.c + Lzma2Enc_Encode2() function worked incorretly, + if (inStream == NULL) and the number of block threads is more than 1. + + +18.03 beta 2018-03-04 +------------------------- +- Asm\x86\LzmaDecOpt.asm: new optimized LZMA decoder written in asm + for x64 with about 30% higher speed than main version of LZMA decoder written in C. +- The speed for single-thread LZMA/LZMA2 decoder written in C was increased by 3%. +- 7-Zip now can use multi-threading for 7z/LZMA2 decoding, + if there are multiple independent data chunks in LZMA2 stream. +- 7-Zip now can use multi-threading for xz decoding, + if there are multiple blocks in xz stream. + + +18.01 2019-01-28 +------------------------- +- The BUG in 17.01 - 18.00 beta was fixed: + XzDec.c : random block unpacking and XzUnpacker_IsBlockFinished() + didn't work correctly for xz archives without checksum (CRC). + + +18.00 beta 2019-01-10 +------------------------- +- The BUG in xz encoder was fixed: + There was memory leak of 16 KB for each file compressed with + xz compression method, if additional filter was used. + + +17.01 beta 2017-08-28 +------------------------- +- Minor speed optimization for LZMA2 (xz and 7z) multi-threading compression. + 7-Zip now uses additional memory buffers for multi-block LZMA2 compression. + CPU utilization was slightly improved. +- 7-zip now creates multi-block xz archives by default. Block size can be + specified with -ms[Size]{m|g} switch. +- xz decoder now can unpack random block from multi-block xz archives. +- 7-Zip command line: @listfile now doesn't work after -- switch. + Use -i@listfile before -- switch instead. +- The BUGs were fixed: + 7-Zip 17.00 beta crashed for commands that write anti-item to 7z archive. + + +17.00 beta 2017-04-29 +------------------------- +- NewHandler.h / NewHandler.cpp: + now it redefines operator new() only for old MSVC compilers (_MSC_VER < 1900). +- C/7zTypes.h : the names of variables in interface structures were changed (vt). +- Some bugs were fixed. 7-Zip could crash in some cases. +- Some internal changes in code. + + +16.04 2016-10-04 +------------------------- +- The bug was fixed in DllSecur.c. + + +16.03 2016-09-28 +------------------------- +- SFX modules now use some protection against DLL preloading attack. +- Some bugs in 7z code were fixed. + + +16.02 2016-05-21 +------------------------- +- The BUG in 16.00 - 16.01 was fixed: + Split Handler (SplitHandler.cpp) returned incorrect + total size value (kpidSize) for split archives. + + +16.01 2016-05-19 +------------------------- +- Some internal changes to reduce the number of compiler warnings. + + +16.00 2016-05-10 +------------------------- +- Some bugs were fixed. + + +15.12 2015-11-19 +------------------------- +- The BUG in C version of 7z decoder was fixed: + 7zDec.c : SzDecodeLzma2() + 7z decoder could mistakenly report about decoding error for some 7z archives + that use LZMA2 compression method. + The probability to get that mistaken decoding error report was about + one error per 16384 solid blocks for solid blocks larger than 16 KB (compressed size). +- The BUG (in 9.26-15.11) in C version of 7z decoder was fixed: + 7zArcIn.c : SzReadHeader2() + 7z decoder worked incorrectly for 7z archives that contain + empty solid blocks, that can be placed to 7z archive, if some file is + unavailable for reading during archive creation. + + +15.09 beta 2015-10-16 +------------------------- +- The BUG in LZMA / LZMA2 encoding code was fixed. + The BUG in LzFind.c::MatchFinder_ReadBlock() function. + If input data size is larger than (4 GiB - dictionary_size), + the following code worked incorrectly: + - LZMA : LzmaEnc_MemEncode(), LzmaEncode() : LZMA encoding functions + for compressing from memory to memory. + That BUG is not related to LZMA encoder version that works via streams. + - LZMA2 : multi-threaded version of LZMA2 encoder worked incorrectly, if + default value of chunk size (CLzma2EncProps::blockSize) is changed + to value larger than (4 GiB - dictionary_size). + + +9.38 beta 2015-01-03 +------------------------- +- The BUG in 9.31-9.37 was fixed: + IArchiveGetRawProps interface was disabled for 7z archives. +- The BUG in 9.26-9.36 was fixed: + Some code in CPP\7zip\Archive\7z\ worked correctly only under Windows. + + +9.36 beta 2014-12-26 +------------------------- +- The BUG in command line version was fixed: + 7-Zip created temporary archive in current folder during update archive + operation, if -w{Path} switch was not specified. + The fixed 7-Zip creates temporary archive in folder that contains updated archive. +- The BUG in 9.33-9.35 was fixed: + 7-Zip silently ignored file reading errors during 7z or gz archive creation, + and the created archive contained only part of file that was read before error. + The fixed 7-Zip stops archive creation and it reports about error. + + +9.35 beta 2014-12-07 +------------------------- +- 7zr.exe now support AES encryption. +- SFX mudules were added to LZMA SDK +- Some bugs were fixed. + + +9.21 beta 2011-04-11 +------------------------- +- New class FString for file names at file systems. +- Speed optimization in CRC code for big-endian CPUs. +- The BUG in Lzma2Dec.c was fixed: + Lzma2Decode function didn't work. + + +9.18 beta 2010-11-02 +------------------------- +- New small SFX module for installers (SfxSetup). + + +9.12 beta 2010-03-24 +------------------------- +- The BUG in LZMA SDK 9.* was fixed: LZMA2 codec didn't work, + if more than 10 threads were used (or more than 20 threads in some modes). + + +9.11 beta 2010-03-15 +------------------------- +- PPMd compression method support + + +9.09 2009-12-12 +------------------------- +- The bug was fixed: + Utf16_To_Utf8 funstions in UTFConvert.cpp and 7zMain.c + incorrectly converted surrogate characters (the code >= 0x10000) to UTF-8. +- Some bugs were fixed + + +9.06 2009-08-17 +------------------------- +- Some changes in ANSI-C 7z Decoder interfaces. + + +9.04 2009-05-30 +------------------------- +- LZMA2 compression method support +- xz format support + + +4.65 2009-02-03 +------------------------- +- Some minor fixes + + +4.63 2008-12-31 +------------------------- +- Some minor fixes + + +4.61 beta 2008-11-23 +------------------------- +- The bug in ANSI-C LZMA Decoder was fixed: + If encoded stream was corrupted, decoder could access memory + outside of allocated range. +- Some changes in ANSI-C 7z Decoder interfaces. +- LZMA SDK is placed in the public domain. + + +4.60 beta 2008-08-19 +------------------------- +- Some minor fixes. + + +4.59 beta 2008-08-13 +------------------------- +- The bug was fixed: + LZMA Encoder in fast compression mode could access memory outside of + allocated range in some rare cases. + + +4.58 beta 2008-05-05 +------------------------- +- ANSI-C LZMA Decoder was rewritten for speed optimizations. +- ANSI-C LZMA Encoder was included to LZMA SDK. +- C++ LZMA code now is just wrapper over ANSI-C code. + + +4.57 2007-12-12 +------------------------- +- Speed optimizations in Ñ++ LZMA Decoder. +- Small changes for more compatibility with some C/C++ compilers. + + +4.49 beta 2007-07-05 +------------------------- +- .7z ANSI-C Decoder: + - now it supports BCJ and BCJ2 filters + - now it supports files larger than 4 GB. + - now it supports "Last Write Time" field for files. +- C++ code for .7z archives compressing/decompressing from 7-zip + was included to LZMA SDK. + + +4.43 2006-06-04 +------------------------- +- Small changes for more compatibility with some C/C++ compilers. + + +4.42 2006-05-15 +------------------------- +- Small changes in .h files in ANSI-C version. + + +4.39 beta 2006-04-14 +------------------------- +- The bug in versions 4.33b:4.38b was fixed: + C++ version of LZMA encoder could not correctly compress + files larger than 2 GB with HC4 match finder (-mfhc4). + + +4.37 beta 2005-04-06 +------------------------- +- Fixes in C++ code: code could no be compiled if _NO_EXCEPTIONS was defined. + + +4.35 beta 2005-03-02 +------------------------- +- The bug was fixed in C++ version of LZMA Decoder: + If encoded stream was corrupted, decoder could access memory + outside of allocated range. + + +4.34 beta 2006-02-27 +------------------------- +- Compressing speed and memory requirements for compressing were increased +- LZMA now can use only these match finders: HC4, BT2, BT3, BT4 + + +4.32 2005-12-09 +------------------------- +- Java version of LZMA SDK was included + + +4.30 2005-11-20 +------------------------- +- Compression ratio was improved in -a2 mode +- Speed optimizations for compressing in -a2 mode +- -fb switch now supports values up to 273 +- The bug in 7z_C (7zIn.c) was fixed: + It used Alloc/Free functions from different memory pools. + So if program used two memory pools, it worked incorrectly. +- 7z_C: .7z format supporting was improved +- LZMA# SDK (C#.NET version) was included + + +4.27 (Updated) 2005-09-21 +------------------------- +- Some GUIDs/interfaces in C++ were changed. + IStream.h: + ISequentialInStream::Read now works as old ReadPart + ISequentialOutStream::Write now works as old WritePart + + +4.27 2005-08-07 +------------------------- +- The bug in LzmaDecodeSize.c was fixed: + if _LZMA_IN_CB and _LZMA_OUT_READ were defined, + decompressing worked incorrectly. + + +4.26 2005-08-05 +------------------------- +- Fixes in 7z_C code and LzmaTest.c: + previous versions could work incorrectly, + if malloc(0) returns 0 + + +4.23 2005-06-29 +------------------------- +- Small fixes in C++ code + + +4.22 2005-06-10 +------------------------- +- Small fixes + + +4.21 2005-06-08 +------------------------- +- Interfaces for ANSI-C LZMA Decoder (LzmaDecode.c) were changed +- New additional version of ANSI-C LZMA Decoder with zlib-like interface: + - LzmaStateDecode.h + - LzmaStateDecode.c + - LzmaStateTest.c +- ANSI-C LZMA Decoder now can decompress files larger than 4 GB + + +4.17 2005-04-18 +------------------------- +- New example for RAM->RAM compressing/decompressing: + LZMA + BCJ (filter for x86 code): + - LzmaRam.h + - LzmaRam.cpp + - LzmaRamDecode.h + - LzmaRamDecode.c + - -f86 switch for lzma.exe + + +4.16 2005-03-29 +------------------------- +- The bug was fixed in LzmaDecode.c (ANSI-C LZMA Decoder): + If _LZMA_OUT_READ was defined, and if encoded stream was corrupted, + decoder could access memory outside of allocated range. +- Speed optimization of ANSI-C LZMA Decoder (now it's about 20% faster). + Old version of LZMA Decoder now is in file LzmaDecodeSize.c. + LzmaDecodeSize.c can provide slightly smaller code than LzmaDecode.c +- Small speed optimization in LZMA C++ code +- filter for SPARC's code was added +- Simplified version of .7z ANSI-C Decoder was included + + +4.06 2004-09-05 +------------------------- +- The bug in v4.05 was fixed: + LZMA-Encoder didn't release output stream in some cases. + + +4.05 2004-08-25 +------------------------- +- Source code of filters for x86, IA-64, ARM, ARM-Thumb + and PowerPC code was included to SDK +- Some internal minor changes + + +4.04 2004-07-28 +------------------------- +- More compatibility with some C++ compilers + + +4.03 2004-06-18 +------------------------- +- "Benchmark" command was added. It measures compressing + and decompressing speed and shows rating values. + Also it checks hardware errors. + + +4.02 2004-06-10 +------------------------- +- C++ LZMA Encoder/Decoder code now is more portable + and it can be compiled by GCC on Linux. + + +4.01 2004-02-15 +------------------------- +- Some detection of data corruption was enabled. + LzmaDecode.c / RangeDecoderReadByte + ..... + { + rd->ExtraBytes = 1; + return 0xFF; + } + + +4.00 2004-02-13 +------------------------- +- Original version of LZMA SDK + + + +HISTORY of the LZMA +------------------- + 2001-2008: Improvements to LZMA compressing/decompressing code, + keeping compatibility with original LZMA format + 1996-2001: Development of LZMA compression format + + Some milestones: + + 2001-08-30: LZMA compression was added to 7-Zip + 1999-01-02: First version of 7-Zip was released + + +End of document diff --git a/deps/LZMA-SDK/DOC/lzma-sdk.txt b/deps/LZMA-SDK/DOC/lzma-sdk.txt index d54aad328..ed1a431ee 100644 --- a/deps/LZMA-SDK/DOC/lzma-sdk.txt +++ b/deps/LZMA-SDK/DOC/lzma-sdk.txt @@ -1,404 +1,404 @@ -LZMA SDK 22.01 --------------- - -LZMA SDK provides the documentation, samples, header files, -libraries, and tools you need to develop applications that -use 7z / LZMA / LZMA2 / XZ compression. - -LZMA is an improved version of famous LZ77 compression algorithm. -It was improved in way of maximum increasing of compression ratio, -keeping high decompression speed and low memory requirements for -decompressing. - -LZMA2 is a LZMA based compression method. LZMA2 provides better -multithreading support for compression than LZMA and some other improvements. - -7z is a file format for data compression and file archiving. -7z is a main file format for 7-Zip compression program (www.7-zip.org). -7z format supports different compression methods: LZMA, LZMA2 and others. -7z also supports AES-256 based encryption. - -XZ is a file format for data compression that uses LZMA2 compression. -XZ format provides additional features: SHA/CRC check, filters for -improved compression ratio, splitting to blocks and streams, - - - -LICENSE -------- - -LZMA SDK is written and placed in the public domain by Igor Pavlov. - -Some code in LZMA SDK is based on public domain code from another developers: - 1) PPMd var.H (2001): Dmitry Shkarin - 2) SHA-256: Wei Dai (Crypto++ library) - -Anyone is free to copy, modify, publish, use, compile, sell, or distribute the -original LZMA SDK code, either in source code form or as a compiled binary, for -any purpose, commercial or non-commercial, and by any means. - -LZMA SDK code is compatible with open source licenses, for example, you can -include it to GNU GPL or GNU LGPL code. - - -LZMA SDK Contents ------------------ - - Source code: - - - C / C++ / C# / Java - LZMA compression and decompression - - C / C++ - LZMA2 compression and decompression - - C / C++ - XZ compression and decompression - - C - 7z decompression - - C++ - 7z compression and decompression - - C - small SFXs for installers (7z decompression) - - C++ - SFXs and SFXs for installers (7z decompression) - - Precomiled binaries: - - - console programs for lzma / 7z / xz compression and decompression - - SFX modules for installers. - - -UNIX/Linux version ------------------- -There are several otpions to compile 7-Zip with different compilers: gcc and clang. -Also 7-Zip code contains two versions for some critical parts of code: in C and in Assembeler. -So if you compile the version with Assembeler code, you will get faster 7-Zip binary. - -7-Zip's assembler code uses the following syntax for different platforms: - -1) x86 and x86-64 (AMD64): MASM syntax. - There are 2 programs that supports MASM syntax in Linux. -' 'Asmc Macro Assembler and JWasm. But JWasm now doesn't support some - cpu instructions used in 7-Zip. - So you must install Asmc Macro Assembler in Linux, if you want to compile fastest version - of 7-Zip x86 and x86-64: - https://github.com/nidud/asmc - -2) arm64: GNU assembler for ARM64 with preprocessor. - That systax of that arm64 assembler code in 7-Zip is supported by GCC and CLANG for ARM64. - -There are different binaries that can be compiled from 7-Zip source. -There are 2 main files in folder for compiling: - makefile - that can be used for compiling Windows version of 7-Zip with nmake command - makefile.gcc - that can be used for compiling Linux/macOS versions of 7-Zip with make command - -At first you must change the current folder to folder that contains `makefile.gcc`: - - cd CPP/7zip/Bundles/Alone7z - -Then you can compile `makefile.gcc` with the command: - - make -j -f makefile.gcc - -Also there are additional "*.mak" files in folder "CPP/7zip/" that can be used to compile -7-Zip binaries with optimized code and optimzing options. - -To compile with GCC without assembler: - cd CPP/7zip/Bundles/Alone7z - make -j -f ../../cmpl_gcc.mak - -To compile with CLANG without assembler: - make -j -f ../../cmpl_clang.mak - -To compile 7-Zip for x86-64 with asmc assembler: - make -j -f ../../cmpl_gcc_x64.mak - -To compile 7-Zip for arm64 with assembler: - make -j -f ../../cmpl_gcc_arm64.mak - -To compile 7-Zip for arm64 for macOS: - make -j -f ../../cmpl_mac_arm64.mak - -Also you can change some compiler options in the mak files: - cmpl_gcc.mak - var_gcc.mak - warn_gcc.mak - - - -Also you can use p7zip (port of 7-Zip for POSIX systems like Unix or Linux): - - http://p7zip.sourceforge.net/ - - -Files ------ - -DOC/7zC.txt - 7z ANSI-C Decoder description -DOC/7zFormat.txt - 7z Format description -DOC/installer.txt - information about 7-Zip for installers -DOC/lzma.txt - LZMA compression description -DOC/lzma-sdk.txt - LZMA SDK description (this file) -DOC/lzma-history.txt - history of LZMA SDK -DOC/lzma-specification.txt - Specification of LZMA -DOC/Methods.txt - Compression method IDs for .7z - -bin/installer/ - example script to create installer that uses SFX module, - -bin/7zdec.exe - simplified 7z archive decoder -bin/7zr.exe - 7-Zip console program (reduced version) -bin/x64/7zr.exe - 7-Zip console program (reduced version) (x64 version) -bin/lzma.exe - file->file LZMA encoder/decoder for Windows -bin/7zS2.sfx - small SFX module for installers (GUI version) -bin/7zS2con.sfx - small SFX module for installers (Console version) -bin/7zSD.sfx - SFX module for installers. - - -7zDec.exe ---------- -7zDec.exe is simplified 7z archive decoder. -It supports only LZMA, LZMA2, and PPMd methods. -7zDec decodes whole solid block from 7z archive to RAM. -The RAM consumption can be high. - - - - -Source code structure ---------------------- - - -Asm/ - asm files (optimized code for CRC calculation and Intel-AES encryption) - -C/ - C files (compression / decompression and other) - Util/ - 7z - 7z decoder program (decoding 7z files) - Lzma - LZMA program (file->file LZMA encoder/decoder). - LzmaLib - LZMA library (.DLL for Windows) - SfxSetup - small SFX module for installers - -CPP/ -- CPP files - - Common - common files for C++ projects - Windows - common files for Windows related code - - 7zip - files related to 7-Zip - - Archive - files related to archiving - - Common - common files for archive handling - 7z - 7z C++ Encoder/Decoder - - Bundles - Modules that are bundles of other modules (files) - - Alone7z - 7zr.exe: Standalone 7-Zip console program (reduced version) - Format7zExtractR - 7zxr.dll: Reduced version of 7z DLL: extracting from 7z/LZMA/BCJ/BCJ2. - Format7zR - 7zr.dll: Reduced version of 7z DLL: extracting/compressing to 7z/LZMA/BCJ/BCJ2 - LzmaCon - lzma.exe: LZMA compression/decompression - LzmaSpec - example code for LZMA Specification - SFXCon - 7zCon.sfx: Console 7z SFX module - SFXSetup - 7zS.sfx: 7z SFX module for installers - SFXWin - 7z.sfx: GUI 7z SFX module - - Common - common files for 7-Zip - - Compress - files for compression/decompression - - Crypto - files for encryption / decompression - - UI - User Interface files - - Client7z - Test application for 7za.dll, 7zr.dll, 7zxr.dll - Common - Common UI files - Console - Code for console program (7z.exe) - Explorer - Some code from 7-Zip Shell extension - FileManager - Some GUI code from 7-Zip File Manager - GUI - Some GUI code from 7-Zip - - -CS/ - C# files - 7zip - Common - some common files for 7-Zip - Compress - files related to compression/decompression - LZ - files related to LZ (Lempel-Ziv) compression algorithm - LZMA - LZMA compression/decompression - LzmaAlone - file->file LZMA compression/decompression - RangeCoder - Range Coder (special code of compression/decompression) - -Java/ - Java files - SevenZip - Compression - files related to compression/decompression - LZ - files related to LZ (Lempel-Ziv) compression algorithm - LZMA - LZMA compression/decompression - RangeCoder - Range Coder (special code of compression/decompression) - - -Note: - Asm / C / C++ source code of LZMA SDK is part of 7-Zip's source code. - 7-Zip's source code can be downloaded from 7-Zip's SourceForge page: - - http://sourceforge.net/projects/sevenzip/ - - - -LZMA features -------------- - - Variable dictionary size (up to 1 GB) - - Estimated compressing speed: about 2 MB/s on 2 GHz CPU - - Estimated decompressing speed: - - 20-30 MB/s on modern 2 GHz cpu - - 1-2 MB/s on 200 MHz simple RISC cpu: (ARM, MIPS, PowerPC) - - Small memory requirements for decompressing (16 KB + DictionarySize) - - Small code size for decompressing: 5-8 KB - -LZMA decoder uses only integer operations and can be -implemented in any modern 32-bit CPU (or on 16-bit CPU with some conditions). - -Some critical operations that affect the speed of LZMA decompression: - 1) 32*16 bit integer multiply - 2) Mispredicted branches (penalty mostly depends from pipeline length) - 3) 32-bit shift and arithmetic operations - -The speed of LZMA decompressing mostly depends from CPU speed. -Memory speed has no big meaning. But if your CPU has small data cache, -overall weight of memory speed will slightly increase. - - -How To Use ----------- - -Using LZMA encoder/decoder executable --------------------------------------- - -Usage: LZMA inputFile outputFile [...] - - e: encode file - - d: decode file - - b: Benchmark. There are two tests: compressing and decompressing - with LZMA method. Benchmark shows rating in MIPS (million - instructions per second). Rating value is calculated from - measured speed and it is normalized with Intel's Core 2 results. - Also Benchmark checks possible hardware errors (RAM - errors in most cases). Benchmark uses these settings: - (-a1, -d21, -fb32, -mfbt4). You can change only -d parameter. - Also you can change the number of iterations. Example for 30 iterations: - LZMA b 30 - Default number of iterations is 10. - - - - - -a{N}: set compression mode 0 = fast, 1 = normal - default: 1 (normal) - - d{N}: Sets Dictionary size - [0, 30], default: 23 (8MB) - The maximum value for dictionary size is 1 GB = 2^30 bytes. - Dictionary size is calculated as DictionarySize = 2^N bytes. - For decompressing file compressed by LZMA method with dictionary - size D = 2^N you need about D bytes of memory (RAM). - - -fb{N}: set number of fast bytes - [5, 273], default: 128 - Usually big number gives a little bit better compression ratio - and slower compression process. - - -lc{N}: set number of literal context bits - [0, 8], default: 3 - Sometimes lc=4 gives gain for big files. - - -lp{N}: set number of literal pos bits - [0, 4], default: 0 - lp switch is intended for periodical data when period is - equal 2^N. For example, for 32-bit (4 bytes) - periodical data you can use lp=2. Often it's better to set lc0, - if you change lp switch. - - -pb{N}: set number of pos bits - [0, 4], default: 2 - pb switch is intended for periodical data - when period is equal 2^N. - - -mf{MF_ID}: set Match Finder. Default: bt4. - Algorithms from hc* group doesn't provide good compression - ratio, but they often works pretty fast in combination with - fast mode (-a0). - - Memory requirements depend from dictionary size - (parameter "d" in table below). - - MF_ID Memory Description - - bt2 d * 9.5 + 4MB Binary Tree with 2 bytes hashing. - bt3 d * 11.5 + 4MB Binary Tree with 3 bytes hashing. - bt4 d * 11.5 + 4MB Binary Tree with 4 bytes hashing. - hc4 d * 7.5 + 4MB Hash Chain with 4 bytes hashing. - - -eos: write End Of Stream marker. By default LZMA doesn't write - eos marker, since LZMA decoder knows uncompressed size - stored in .lzma file header. - - -si: Read data from stdin (it will write End Of Stream marker). - -so: Write data to stdout - - -Examples: - -1) LZMA e file.bin file.lzma -d16 -lc0 - -compresses file.bin to file.lzma with 64 KB dictionary (2^16=64K) -and 0 literal context bits. -lc0 allows to reduce memory requirements -for decompression. - - -2) LZMA e file.bin file.lzma -lc0 -lp2 - -compresses file.bin to file.lzma with settings suitable -for 32-bit periodical data (for example, ARM or MIPS code). - -3) LZMA d file.lzma file.bin - -decompresses file.lzma to file.bin. - - -Compression ratio hints ------------------------ - -Recommendations ---------------- - -To increase the compression ratio for LZMA compressing it's desirable -to have aligned data (if it's possible) and also it's desirable to locate -data in such order, where code is grouped in one place and data is -grouped in other place (it's better than such mixing: code, data, code, -data, ...). - - -Filters -------- -You can increase the compression ratio for some data types, using -special filters before compressing. For example, it's possible to -increase the compression ratio on 5-10% for code for those CPU ISAs: -x86, IA-64, ARM, ARM-Thumb, PowerPC, SPARC. - -You can find C source code of such filters in C/Bra*.* files - -You can check the compression ratio gain of these filters with such -7-Zip commands (example for ARM code): -No filter: - 7z a a1.7z a.bin -m0=lzma - -With filter for little-endian ARM code: - 7z a a2.7z a.bin -m0=arm -m1=lzma - -It works in such manner: -Compressing = Filter_encoding + LZMA_encoding -Decompressing = LZMA_decoding + Filter_decoding - -Compressing and decompressing speed of such filters is very high, -so it will not increase decompressing time too much. -Moreover, it reduces decompression time for LZMA_decoding, -since compression ratio with filtering is higher. - -These filters convert CALL (calling procedure) instructions -from relative offsets to absolute addresses, so such data becomes more -compressible. - -For some ISAs (for example, for MIPS) it's impossible to get gain from such filter. - - - ---- - -http://www.7-zip.org -http://www.7-zip.org/sdk.html -http://www.7-zip.org/support.html +LZMA SDK 22.01 +-------------- + +LZMA SDK provides the documentation, samples, header files, +libraries, and tools you need to develop applications that +use 7z / LZMA / LZMA2 / XZ compression. + +LZMA is an improved version of famous LZ77 compression algorithm. +It was improved in way of maximum increasing of compression ratio, +keeping high decompression speed and low memory requirements for +decompressing. + +LZMA2 is a LZMA based compression method. LZMA2 provides better +multithreading support for compression than LZMA and some other improvements. + +7z is a file format for data compression and file archiving. +7z is a main file format for 7-Zip compression program (www.7-zip.org). +7z format supports different compression methods: LZMA, LZMA2 and others. +7z also supports AES-256 based encryption. + +XZ is a file format for data compression that uses LZMA2 compression. +XZ format provides additional features: SHA/CRC check, filters for +improved compression ratio, splitting to blocks and streams, + + + +LICENSE +------- + +LZMA SDK is written and placed in the public domain by Igor Pavlov. + +Some code in LZMA SDK is based on public domain code from another developers: + 1) PPMd var.H (2001): Dmitry Shkarin + 2) SHA-256: Wei Dai (Crypto++ library) + +Anyone is free to copy, modify, publish, use, compile, sell, or distribute the +original LZMA SDK code, either in source code form or as a compiled binary, for +any purpose, commercial or non-commercial, and by any means. + +LZMA SDK code is compatible with open source licenses, for example, you can +include it to GNU GPL or GNU LGPL code. + + +LZMA SDK Contents +----------------- + + Source code: + + - C / C++ / C# / Java - LZMA compression and decompression + - C / C++ - LZMA2 compression and decompression + - C / C++ - XZ compression and decompression + - C - 7z decompression + - C++ - 7z compression and decompression + - C - small SFXs for installers (7z decompression) + - C++ - SFXs and SFXs for installers (7z decompression) + + Precomiled binaries: + + - console programs for lzma / 7z / xz compression and decompression + - SFX modules for installers. + + +UNIX/Linux version +------------------ +There are several otpions to compile 7-Zip with different compilers: gcc and clang. +Also 7-Zip code contains two versions for some critical parts of code: in C and in Assembeler. +So if you compile the version with Assembeler code, you will get faster 7-Zip binary. + +7-Zip's assembler code uses the following syntax for different platforms: + +1) x86 and x86-64 (AMD64): MASM syntax. + There are 2 programs that supports MASM syntax in Linux. +' 'Asmc Macro Assembler and JWasm. But JWasm now doesn't support some + cpu instructions used in 7-Zip. + So you must install Asmc Macro Assembler in Linux, if you want to compile fastest version + of 7-Zip x86 and x86-64: + https://github.com/nidud/asmc + +2) arm64: GNU assembler for ARM64 with preprocessor. + That systax of that arm64 assembler code in 7-Zip is supported by GCC and CLANG for ARM64. + +There are different binaries that can be compiled from 7-Zip source. +There are 2 main files in folder for compiling: + makefile - that can be used for compiling Windows version of 7-Zip with nmake command + makefile.gcc - that can be used for compiling Linux/macOS versions of 7-Zip with make command + +At first you must change the current folder to folder that contains `makefile.gcc`: + + cd CPP/7zip/Bundles/Alone7z + +Then you can compile `makefile.gcc` with the command: + + make -j -f makefile.gcc + +Also there are additional "*.mak" files in folder "CPP/7zip/" that can be used to compile +7-Zip binaries with optimized code and optimzing options. + +To compile with GCC without assembler: + cd CPP/7zip/Bundles/Alone7z + make -j -f ../../cmpl_gcc.mak + +To compile with CLANG without assembler: + make -j -f ../../cmpl_clang.mak + +To compile 7-Zip for x86-64 with asmc assembler: + make -j -f ../../cmpl_gcc_x64.mak + +To compile 7-Zip for arm64 with assembler: + make -j -f ../../cmpl_gcc_arm64.mak + +To compile 7-Zip for arm64 for macOS: + make -j -f ../../cmpl_mac_arm64.mak + +Also you can change some compiler options in the mak files: + cmpl_gcc.mak + var_gcc.mak + warn_gcc.mak + + + +Also you can use p7zip (port of 7-Zip for POSIX systems like Unix or Linux): + + http://p7zip.sourceforge.net/ + + +Files +----- + +DOC/7zC.txt - 7z ANSI-C Decoder description +DOC/7zFormat.txt - 7z Format description +DOC/installer.txt - information about 7-Zip for installers +DOC/lzma.txt - LZMA compression description +DOC/lzma-sdk.txt - LZMA SDK description (this file) +DOC/lzma-history.txt - history of LZMA SDK +DOC/lzma-specification.txt - Specification of LZMA +DOC/Methods.txt - Compression method IDs for .7z + +bin/installer/ - example script to create installer that uses SFX module, + +bin/7zdec.exe - simplified 7z archive decoder +bin/7zr.exe - 7-Zip console program (reduced version) +bin/x64/7zr.exe - 7-Zip console program (reduced version) (x64 version) +bin/lzma.exe - file->file LZMA encoder/decoder for Windows +bin/7zS2.sfx - small SFX module for installers (GUI version) +bin/7zS2con.sfx - small SFX module for installers (Console version) +bin/7zSD.sfx - SFX module for installers. + + +7zDec.exe +--------- +7zDec.exe is simplified 7z archive decoder. +It supports only LZMA, LZMA2, and PPMd methods. +7zDec decodes whole solid block from 7z archive to RAM. +The RAM consumption can be high. + + + + +Source code structure +--------------------- + + +Asm/ - asm files (optimized code for CRC calculation and Intel-AES encryption) + +C/ - C files (compression / decompression and other) + Util/ + 7z - 7z decoder program (decoding 7z files) + Lzma - LZMA program (file->file LZMA encoder/decoder). + LzmaLib - LZMA library (.DLL for Windows) + SfxSetup - small SFX module for installers + +CPP/ -- CPP files + + Common - common files for C++ projects + Windows - common files for Windows related code + + 7zip - files related to 7-Zip + + Archive - files related to archiving + + Common - common files for archive handling + 7z - 7z C++ Encoder/Decoder + + Bundles - Modules that are bundles of other modules (files) + + Alone7z - 7zr.exe: Standalone 7-Zip console program (reduced version) + Format7zExtractR - 7zxr.dll: Reduced version of 7z DLL: extracting from 7z/LZMA/BCJ/BCJ2. + Format7zR - 7zr.dll: Reduced version of 7z DLL: extracting/compressing to 7z/LZMA/BCJ/BCJ2 + LzmaCon - lzma.exe: LZMA compression/decompression + LzmaSpec - example code for LZMA Specification + SFXCon - 7zCon.sfx: Console 7z SFX module + SFXSetup - 7zS.sfx: 7z SFX module for installers + SFXWin - 7z.sfx: GUI 7z SFX module + + Common - common files for 7-Zip + + Compress - files for compression/decompression + + Crypto - files for encryption / decompression + + UI - User Interface files + + Client7z - Test application for 7za.dll, 7zr.dll, 7zxr.dll + Common - Common UI files + Console - Code for console program (7z.exe) + Explorer - Some code from 7-Zip Shell extension + FileManager - Some GUI code from 7-Zip File Manager + GUI - Some GUI code from 7-Zip + + +CS/ - C# files + 7zip + Common - some common files for 7-Zip + Compress - files related to compression/decompression + LZ - files related to LZ (Lempel-Ziv) compression algorithm + LZMA - LZMA compression/decompression + LzmaAlone - file->file LZMA compression/decompression + RangeCoder - Range Coder (special code of compression/decompression) + +Java/ - Java files + SevenZip + Compression - files related to compression/decompression + LZ - files related to LZ (Lempel-Ziv) compression algorithm + LZMA - LZMA compression/decompression + RangeCoder - Range Coder (special code of compression/decompression) + + +Note: + Asm / C / C++ source code of LZMA SDK is part of 7-Zip's source code. + 7-Zip's source code can be downloaded from 7-Zip's SourceForge page: + + http://sourceforge.net/projects/sevenzip/ + + + +LZMA features +------------- + - Variable dictionary size (up to 1 GB) + - Estimated compressing speed: about 2 MB/s on 2 GHz CPU + - Estimated decompressing speed: + - 20-30 MB/s on modern 2 GHz cpu + - 1-2 MB/s on 200 MHz simple RISC cpu: (ARM, MIPS, PowerPC) + - Small memory requirements for decompressing (16 KB + DictionarySize) + - Small code size for decompressing: 5-8 KB + +LZMA decoder uses only integer operations and can be +implemented in any modern 32-bit CPU (or on 16-bit CPU with some conditions). + +Some critical operations that affect the speed of LZMA decompression: + 1) 32*16 bit integer multiply + 2) Mispredicted branches (penalty mostly depends from pipeline length) + 3) 32-bit shift and arithmetic operations + +The speed of LZMA decompressing mostly depends from CPU speed. +Memory speed has no big meaning. But if your CPU has small data cache, +overall weight of memory speed will slightly increase. + + +How To Use +---------- + +Using LZMA encoder/decoder executable +-------------------------------------- + +Usage: LZMA inputFile outputFile [...] + + e: encode file + + d: decode file + + b: Benchmark. There are two tests: compressing and decompressing + with LZMA method. Benchmark shows rating in MIPS (million + instructions per second). Rating value is calculated from + measured speed and it is normalized with Intel's Core 2 results. + Also Benchmark checks possible hardware errors (RAM + errors in most cases). Benchmark uses these settings: + (-a1, -d21, -fb32, -mfbt4). You can change only -d parameter. + Also you can change the number of iterations. Example for 30 iterations: + LZMA b 30 + Default number of iterations is 10. + + + + + -a{N}: set compression mode 0 = fast, 1 = normal + default: 1 (normal) + + d{N}: Sets Dictionary size - [0, 30], default: 23 (8MB) + The maximum value for dictionary size is 1 GB = 2^30 bytes. + Dictionary size is calculated as DictionarySize = 2^N bytes. + For decompressing file compressed by LZMA method with dictionary + size D = 2^N you need about D bytes of memory (RAM). + + -fb{N}: set number of fast bytes - [5, 273], default: 128 + Usually big number gives a little bit better compression ratio + and slower compression process. + + -lc{N}: set number of literal context bits - [0, 8], default: 3 + Sometimes lc=4 gives gain for big files. + + -lp{N}: set number of literal pos bits - [0, 4], default: 0 + lp switch is intended for periodical data when period is + equal 2^N. For example, for 32-bit (4 bytes) + periodical data you can use lp=2. Often it's better to set lc0, + if you change lp switch. + + -pb{N}: set number of pos bits - [0, 4], default: 2 + pb switch is intended for periodical data + when period is equal 2^N. + + -mf{MF_ID}: set Match Finder. Default: bt4. + Algorithms from hc* group doesn't provide good compression + ratio, but they often works pretty fast in combination with + fast mode (-a0). + + Memory requirements depend from dictionary size + (parameter "d" in table below). + + MF_ID Memory Description + + bt2 d * 9.5 + 4MB Binary Tree with 2 bytes hashing. + bt3 d * 11.5 + 4MB Binary Tree with 3 bytes hashing. + bt4 d * 11.5 + 4MB Binary Tree with 4 bytes hashing. + hc4 d * 7.5 + 4MB Hash Chain with 4 bytes hashing. + + -eos: write End Of Stream marker. By default LZMA doesn't write + eos marker, since LZMA decoder knows uncompressed size + stored in .lzma file header. + + -si: Read data from stdin (it will write End Of Stream marker). + -so: Write data to stdout + + +Examples: + +1) LZMA e file.bin file.lzma -d16 -lc0 + +compresses file.bin to file.lzma with 64 KB dictionary (2^16=64K) +and 0 literal context bits. -lc0 allows to reduce memory requirements +for decompression. + + +2) LZMA e file.bin file.lzma -lc0 -lp2 + +compresses file.bin to file.lzma with settings suitable +for 32-bit periodical data (for example, ARM or MIPS code). + +3) LZMA d file.lzma file.bin + +decompresses file.lzma to file.bin. + + +Compression ratio hints +----------------------- + +Recommendations +--------------- + +To increase the compression ratio for LZMA compressing it's desirable +to have aligned data (if it's possible) and also it's desirable to locate +data in such order, where code is grouped in one place and data is +grouped in other place (it's better than such mixing: code, data, code, +data, ...). + + +Filters +------- +You can increase the compression ratio for some data types, using +special filters before compressing. For example, it's possible to +increase the compression ratio on 5-10% for code for those CPU ISAs: +x86, IA-64, ARM, ARM-Thumb, PowerPC, SPARC. + +You can find C source code of such filters in C/Bra*.* files + +You can check the compression ratio gain of these filters with such +7-Zip commands (example for ARM code): +No filter: + 7z a a1.7z a.bin -m0=lzma + +With filter for little-endian ARM code: + 7z a a2.7z a.bin -m0=arm -m1=lzma + +It works in such manner: +Compressing = Filter_encoding + LZMA_encoding +Decompressing = LZMA_decoding + Filter_decoding + +Compressing and decompressing speed of such filters is very high, +so it will not increase decompressing time too much. +Moreover, it reduces decompression time for LZMA_decoding, +since compression ratio with filtering is higher. + +These filters convert CALL (calling procedure) instructions +from relative offsets to absolute addresses, so such data becomes more +compressible. + +For some ISAs (for example, for MIPS) it's impossible to get gain from such filter. + + + +--- + +http://www.7-zip.org +http://www.7-zip.org/sdk.html +http://www.7-zip.org/support.html diff --git a/deps/LZMA-SDK/DOC/lzma-specification.txt b/deps/LZMA-SDK/DOC/lzma-specification.txt index b6796df75..ac0cce76a 100644 --- a/deps/LZMA-SDK/DOC/lzma-specification.txt +++ b/deps/LZMA-SDK/DOC/lzma-specification.txt @@ -1,1176 +1,1176 @@ -LZMA specification (DRAFT version) ----------------------------------- - -Author: Igor Pavlov -Date: 2015-06-14 - -This specification defines the format of LZMA compressed data and lzma file format. - -Notation --------- - -We use the syntax of C++ programming language. -We use the following types in C++ code: - unsigned - unsigned integer, at least 16 bits in size - int - signed integer, at least 16 bits in size - UInt64 - 64-bit unsigned integer - UInt32 - 32-bit unsigned integer - UInt16 - 16-bit unsigned integer - Byte - 8-bit unsigned integer - bool - boolean type with two possible values: false, true - - -lzma file format -================ - -The lzma file contains the raw LZMA stream and the header with related properties. - -The files in that format use ".lzma" extension. - -The lzma file format layout: - -Offset Size Description - - 0 1 LZMA model properties (lc, lp, pb) in encoded form - 1 4 Dictionary size (32-bit unsigned integer, little-endian) - 5 8 Uncompressed size (64-bit unsigned integer, little-endian) - 13 Compressed data (LZMA stream) - -LZMA properties: - - name Range Description - - lc [0, 8] the number of "literal context" bits - lp [0, 4] the number of "literal pos" bits - pb [0, 4] the number of "pos" bits -dictSize [0, 2^32 - 1] the dictionary size - -The following code encodes LZMA properties: - -void EncodeProperties(Byte *properties) -{ - properties[0] = (Byte)((pb * 5 + lp) * 9 + lc); - Set_UInt32_LittleEndian(properties + 1, dictSize); -} - -If the value of dictionary size in properties is smaller than (1 << 12), -the LZMA decoder must set the dictionary size variable to (1 << 12). - -#define LZMA_DIC_MIN (1 << 12) - - unsigned lc, pb, lp; - UInt32 dictSize; - UInt32 dictSizeInProperties; - - void DecodeProperties(const Byte *properties) - { - unsigned d = properties[0]; - if (d >= (9 * 5 * 5)) - throw "Incorrect LZMA properties"; - lc = d % 9; - d /= 9; - pb = d / 5; - lp = d % 5; - dictSizeInProperties = 0; - for (int i = 0; i < 4; i++) - dictSizeInProperties |= (UInt32)properties[i + 1] << (8 * i); - dictSize = dictSizeInProperties; - if (dictSize < LZMA_DIC_MIN) - dictSize = LZMA_DIC_MIN; - } - -If "Uncompressed size" field contains ones in all 64 bits, it means that -uncompressed size is unknown and there is the "end marker" in stream, -that indicates the end of decoding point. -In opposite case, if the value from "Uncompressed size" field is not -equal to ((2^64) - 1), the LZMA stream decoding must be finished after -specified number of bytes (Uncompressed size) is decoded. And if there -is the "end marker", the LZMA decoder must read that marker also. - - -The new scheme to encode LZMA properties ----------------------------------------- - -If LZMA compression is used for some another format, it's recommended to -use a new improved scheme to encode LZMA properties. That new scheme was -used in xz format that uses the LZMA2 compression algorithm. -The LZMA2 is a new compression algorithm that is based on the LZMA algorithm. - -The dictionary size in LZMA2 is encoded with just one byte and LZMA2 supports -only reduced set of dictionary sizes: - (2 << 11), (3 << 11), - (2 << 12), (3 << 12), - ... - (2 << 30), (3 << 30), - (2 << 31) - 1 - -The dictionary size can be extracted from encoded value with the following code: - - dictSize = (p == 40) ? 0xFFFFFFFF : (((UInt32)2 | ((p) & 1)) << ((p) / 2 + 11)); - -Also there is additional limitation (lc + lp <= 4) in LZMA2 for values of -"lc" and "lp" properties: - - if (lc + lp > 4) - throw "Unsupported properties: (lc + lp) > 4"; - -There are some advantages for LZMA decoder with such (lc + lp) value -limitation. It reduces the maximum size of tables allocated by decoder. -And it reduces the complexity of initialization procedure, that can be -important to keep high speed of decoding of big number of small LZMA streams. - -It's recommended to use that limitation (lc + lp <= 4) for any new format -that uses LZMA compression. Note that the combinations of "lc" and "lp" -parameters, where (lc + lp > 4), can provide significant improvement in -compression ratio only in some rare cases. - -The LZMA properties can be encoded into two bytes in new scheme: - -Offset Size Description - - 0 1 The dictionary size encoded with LZMA2 scheme - 1 1 LZMA model properties (lc, lp, pb) in encoded form - - -The RAM usage -============= - -The RAM usage for LZMA decoder is determined by the following parts: - -1) The Sliding Window (from 4 KiB to 4 GiB). -2) The probability model counter arrays (arrays of 16-bit variables). -3) Some additional state variables (about 10 variables of 32-bit integers). - - -The RAM usage for Sliding Window --------------------------------- - -There are two main scenarios of decoding: - -1) The decoding of full stream to one RAM buffer. - - If we decode full LZMA stream to one output buffer in RAM, the decoder - can use that output buffer as sliding window. So the decoder doesn't - need additional buffer allocated for sliding window. - -2) The decoding to some external storage. - - If we decode LZMA stream to external storage, the decoder must allocate - the buffer for sliding window. The size of that buffer must be equal - or larger than the value of dictionary size from properties of LZMA stream. - -In this specification we describe the code for decoding to some external -storage. The optimized version of code for decoding of full stream to one -output RAM buffer can require some minor changes in code. - - -The RAM usage for the probability model counters ------------------------------------------------- - -The size of the probability model counter arrays is calculated with the -following formula: - -size_of_prob_arrays = 1846 + 768 * (1 << (lp + lc)) - -Each probability model counter is 11-bit unsigned integer. -If we use 16-bit integer variables (2-byte integers) for these probability -model counters, the RAM usage required by probability model counter arrays -can be estimated with the following formula: - - RAM = 4 KiB + 1.5 KiB * (1 << (lp + lc)) - -For example, for default LZMA parameters (lp = 0 and lc = 3), the RAM usage is - - RAM_lc3_lp0 = 4 KiB + 1.5 KiB * 8 = 16 KiB - -The maximum RAM state usage is required for decoding the stream with lp = 4 -and lc = 8: - - RAM_lc8_lp4 = 4 KiB + 1.5 KiB * 4096 = 6148 KiB - -If the decoder uses LZMA2's limited property condition -(lc + lp <= 4), the RAM usage will be not larger than - - RAM_lc_lp_4 = 4 KiB + 1.5 KiB * 16 = 28 KiB - - -The RAM usage for encoder -------------------------- - -There are many variants for LZMA encoding code. -These variants have different values for memory consumption. -Note that memory consumption for LZMA Encoder can not be -smaller than memory consumption of LZMA Decoder for same stream. - -The RAM usage required by modern effective implementation of -LZMA Encoder can be estimated with the following formula: - - Encoder_RAM_Usage = 4 MiB + 11 * dictionarySize. - -But there are some modes of the encoder that require less memory. - - -LZMA Decoding -============= - -The LZMA compression algorithm uses LZ-based compression with Sliding Window -and Range Encoding as entropy coding method. - - -Sliding Window --------------- - -LZMA uses Sliding Window compression similar to LZ77 algorithm. - -LZMA stream must be decoded to the sequence that consists -of MATCHES and LITERALS: - - - a LITERAL is a 8-bit character (one byte). - The decoder just puts that LITERAL to the uncompressed stream. - - - a MATCH is a pair of two numbers (DISTANCE-LENGTH pair). - The decoder takes one byte exactly "DISTANCE" characters behind - current position in the uncompressed stream and puts it to - uncompressed stream. The decoder must repeat it "LENGTH" times. - -The "DISTANCE" can not be larger than dictionary size. -And the "DISTANCE" can not be larger than the number of bytes in -the uncompressed stream that were decoded before that match. - -In this specification we use cyclic buffer to implement Sliding Window -for LZMA decoder: - -class COutWindow -{ - Byte *Buf; - UInt32 Pos; - UInt32 Size; - bool IsFull; - -public: - unsigned TotalPos; - COutStream OutStream; - - COutWindow(): Buf(NULL) {} - ~COutWindow() { delete []Buf; } - - void Create(UInt32 dictSize) - { - Buf = new Byte[dictSize]; - Pos = 0; - Size = dictSize; - IsFull = false; - TotalPos = 0; - } - - void PutByte(Byte b) - { - TotalPos++; - Buf[Pos++] = b; - if (Pos == Size) - { - Pos = 0; - IsFull = true; - } - OutStream.WriteByte(b); - } - - Byte GetByte(UInt32 dist) const - { - return Buf[dist <= Pos ? Pos - dist : Size - dist + Pos]; - } - - void CopyMatch(UInt32 dist, unsigned len) - { - for (; len > 0; len--) - PutByte(GetByte(dist)); - } - - bool CheckDistance(UInt32 dist) const - { - return dist <= Pos || IsFull; - } - - bool IsEmpty() const - { - return Pos == 0 && !IsFull; - } -}; - - -In another implementation it's possible to use one buffer that contains -Sliding Window and the whole data stream after uncompressing. - - -Range Decoder -------------- - -LZMA algorithm uses Range Encoding (1) as entropy coding method. - -LZMA stream contains just one very big number in big-endian encoding. -LZMA decoder uses the Range Decoder to extract a sequence of binary -symbols from that big number. - -The state of the Range Decoder: - -struct CRangeDecoder -{ - UInt32 Range; - UInt32 Code; - InputStream *InStream; - - bool Corrupted; -} - -The notes about UInt32 type for the "Range" and "Code" variables: - - It's possible to use 64-bit (unsigned or signed) integer type - for the "Range" and the "Code" variables instead of 32-bit unsigned, - but some additional code must be used to truncate the values to - low 32-bits after some operations. - - If the programming language does not support 32-bit unsigned integer type - (like in case of JAVA language), it's possible to use 32-bit signed integer, - but some code must be changed. For example, it's required to change the code - that uses comparison operations for UInt32 variables in this specification. - -The Range Decoder can be in some states that can be treated as -"Corruption" in LZMA stream. The Range Decoder uses the variable "Corrupted": - - (Corrupted == false), if the Range Decoder has not detected any corruption. - (Corrupted == true), if the Range Decoder has detected some corruption. - -The reference LZMA Decoder ignores the value of the "Corrupted" variable. -So it continues to decode the stream, even if the corruption can be detected -in the Range Decoder. To provide the full compatibility with output of the -reference LZMA Decoder, another LZMA Decoder implementations must also -ignore the value of the "Corrupted" variable. - -The LZMA Encoder is required to create only such LZMA streams, that will not -lead the Range Decoder to states, where the "Corrupted" variable is set to true. - -The Range Decoder reads first 5 bytes from input stream to initialize -the state: - -bool CRangeDecoder::Init() -{ - Corrupted = false; - Range = 0xFFFFFFFF; - Code = 0; - - Byte b = InStream->ReadByte(); - - for (int i = 0; i < 4; i++) - Code = (Code << 8) | InStream->ReadByte(); - - if (b != 0 || Code == Range) - Corrupted = true; - return b == 0; -} - -The LZMA Encoder always writes ZERO in initial byte of compressed stream. -That scheme allows to simplify the code of the Range Encoder in the -LZMA Encoder. If initial byte is not equal to ZERO, the LZMA Decoder must -stop decoding and report error. - -After the last bit of data was decoded by Range Decoder, the value of the -"Code" variable must be equal to 0. The LZMA Decoder must check it by -calling the IsFinishedOK() function: - - bool IsFinishedOK() const { return Code == 0; } - -If there is corruption in data stream, there is big probability that -the "Code" value will be not equal to 0 in the Finish() function. So that -check in the IsFinishedOK() function provides very good feature for -corruption detection. - -The value of the "Range" variable before each bit decoding can not be smaller -than ((UInt32)1 << 24). The Normalize() function keeps the "Range" value in -described range. - -#define kTopValue ((UInt32)1 << 24) - -void CRangeDecoder::Normalize() -{ - if (Range < kTopValue) - { - Range <<= 8; - Code = (Code << 8) | InStream->ReadByte(); - } -} - -Notes: if the size of the "Code" variable is larger than 32 bits, it's -required to keep only low 32 bits of the "Code" variable after the change -in Normalize() function. - -If the LZMA Stream is not corrupted, the value of the "Code" variable is -always smaller than value of the "Range" variable. -But the Range Decoder ignores some types of corruptions, so the value of -the "Code" variable can be equal or larger than value of the "Range" variable -for some "Corrupted" archives. - - -LZMA uses Range Encoding only with binary symbols of two types: - 1) binary symbols with fixed and equal probabilities (direct bits) - 2) binary symbols with predicted probabilities - -The DecodeDirectBits() function decodes the sequence of direct bits: - -UInt32 CRangeDecoder::DecodeDirectBits(unsigned numBits) -{ - UInt32 res = 0; - do - { - Range >>= 1; - Code -= Range; - UInt32 t = 0 - ((UInt32)Code >> 31); - Code += Range & t; - - if (Code == Range) - Corrupted = true; - - Normalize(); - res <<= 1; - res += t + 1; - } - while (--numBits); - return res; -} - - -The Bit Decoding with Probability Model ---------------------------------------- - -The task of Bit Probability Model is to estimate probabilities of binary -symbols. And then it provides the Range Decoder with that information. -The better prediction provides better compression ratio. -The Bit Probability Model uses statistical data of previous decoded -symbols. - -That estimated probability is presented as 11-bit unsigned integer value -that represents the probability of symbol "0". - -#define kNumBitModelTotalBits 11 - -Mathematical probabilities can be presented with the following formulas: - probability(symbol_0) = prob / 2048. - probability(symbol_1) = 1 - Probability(symbol_0) = - = 1 - prob / 2048 = - = (2048 - prob) / 2048 -where the "prob" variable contains 11-bit integer probability counter. - -It's recommended to use 16-bit unsigned integer type, to store these 11-bit -probability values: - -typedef UInt16 CProb; - -Each probability value must be initialized with value ((1 << 11) / 2), -that represents the state, where probabilities of symbols 0 and 1 -are equal to 0.5: - -#define PROB_INIT_VAL ((1 << kNumBitModelTotalBits) / 2) - -The INIT_PROBS macro is used to initialize the array of CProb variables: - -#define INIT_PROBS(p) \ - { for (unsigned i = 0; i < sizeof(p) / sizeof(p[0]); i++) p[i] = PROB_INIT_VAL; } - - -The DecodeBit() function decodes one bit. -The LZMA decoder provides the pointer to CProb variable that contains -information about estimated probability for symbol 0 and the Range Decoder -updates that CProb variable after decoding. The Range Decoder increases -estimated probability of the symbol that was decoded: - -#define kNumMoveBits 5 - -unsigned CRangeDecoder::DecodeBit(CProb *prob) -{ - unsigned v = *prob; - UInt32 bound = (Range >> kNumBitModelTotalBits) * v; - unsigned symbol; - if (Code < bound) - { - v += ((1 << kNumBitModelTotalBits) - v) >> kNumMoveBits; - Range = bound; - symbol = 0; - } - else - { - v -= v >> kNumMoveBits; - Code -= bound; - Range -= bound; - symbol = 1; - } - *prob = (CProb)v; - Normalize(); - return symbol; -} - - -The Binary Tree of bit model counters -------------------------------------- - -LZMA uses a tree of Bit model variables to decode symbol that needs -several bits for storing. There are two versions of such trees in LZMA: - 1) the tree that decodes bits from high bit to low bit (the normal scheme). - 2) the tree that decodes bits from low bit to high bit (the reverse scheme). - -Each binary tree structure supports different size of decoded symbol -(the size of binary sequence that contains value of symbol). -If that size of decoded symbol is "NumBits" bits, the tree structure -uses the array of (2 << NumBits) counters of CProb type. -But only ((2 << NumBits) - 1) items are used by encoder and decoder. -The first item (the item with index equal to 0) in array is unused. -That scheme with unused array's item allows to simplify the code. - -unsigned BitTreeReverseDecode(CProb *probs, unsigned numBits, CRangeDecoder *rc) -{ - unsigned m = 1; - unsigned symbol = 0; - for (unsigned i = 0; i < numBits; i++) - { - unsigned bit = rc->DecodeBit(&probs[m]); - m <<= 1; - m += bit; - symbol |= (bit << i); - } - return symbol; -} - -template -class CBitTreeDecoder -{ - CProb Probs[(unsigned)1 << NumBits]; - -public: - - void Init() - { - INIT_PROBS(Probs); - } - - unsigned Decode(CRangeDecoder *rc) - { - unsigned m = 1; - for (unsigned i = 0; i < NumBits; i++) - m = (m << 1) + rc->DecodeBit(&Probs[m]); - return m - ((unsigned)1 << NumBits); - } - - unsigned ReverseDecode(CRangeDecoder *rc) - { - return BitTreeReverseDecode(Probs, NumBits, rc); - } -}; - - -LZ part of LZMA ---------------- - -LZ part of LZMA describes details about the decoding of MATCHES and LITERALS. - - -The Literal Decoding --------------------- - -The LZMA Decoder uses (1 << (lc + lp)) tables with CProb values, where -each table contains 0x300 CProb values: - - CProb *LitProbs; - - void CreateLiterals() - { - LitProbs = new CProb[(UInt32)0x300 << (lc + lp)]; - } - - void InitLiterals() - { - UInt32 num = (UInt32)0x300 << (lc + lp); - for (UInt32 i = 0; i < num; i++) - LitProbs[i] = PROB_INIT_VAL; - } - -To select the table for decoding it uses the context that consists of -(lc) high bits from previous literal and (lp) low bits from value that -represents current position in outputStream. - -If (State > 7), the Literal Decoder also uses "matchByte" that represents -the byte in OutputStream at position the is the DISTANCE bytes before -current position, where the DISTANCE is the distance in DISTANCE-LENGTH pair -of latest decoded match. - -The following code decodes one literal and puts it to Sliding Window buffer: - - void DecodeLiteral(unsigned state, UInt32 rep0) - { - unsigned prevByte = 0; - if (!OutWindow.IsEmpty()) - prevByte = OutWindow.GetByte(1); - - unsigned symbol = 1; - unsigned litState = ((OutWindow.TotalPos & ((1 << lp) - 1)) << lc) + (prevByte >> (8 - lc)); - CProb *probs = &LitProbs[(UInt32)0x300 * litState]; - - if (state >= 7) - { - unsigned matchByte = OutWindow.GetByte(rep0 + 1); - do - { - unsigned matchBit = (matchByte >> 7) & 1; - matchByte <<= 1; - unsigned bit = RangeDec.DecodeBit(&probs[((1 + matchBit) << 8) + symbol]); - symbol = (symbol << 1) | bit; - if (matchBit != bit) - break; - } - while (symbol < 0x100); - } - while (symbol < 0x100) - symbol = (symbol << 1) | RangeDec.DecodeBit(&probs[symbol]); - OutWindow.PutByte((Byte)(symbol - 0x100)); - } - - -The match length decoding -------------------------- - -The match length decoder returns normalized (zero-based value) -length of match. That value can be converted to real length of the match -with the following code: - -#define kMatchMinLen 2 - - matchLen = len + kMatchMinLen; - -The match length decoder can return the values from 0 to 271. -And the corresponded real match length values can be in the range -from 2 to 273. - -The following scheme is used for the match length encoding: - - Binary encoding Binary Tree structure Zero-based match length - sequence (binary + decimal): - - 0 xxx LowCoder[posState] xxx - 1 0 yyy MidCoder[posState] yyy + 8 - 1 1 zzzzzzzz HighCoder zzzzzzzz + 16 - -LZMA uses bit model variable "Choice" to decode the first selection bit. - -If the first selection bit is equal to 0, the decoder uses binary tree - LowCoder[posState] to decode 3-bit zero-based match length (xxx). - -If the first selection bit is equal to 1, the decoder uses bit model - variable "Choice2" to decode the second selection bit. - - If the second selection bit is equal to 0, the decoder uses binary tree - MidCoder[posState] to decode 3-bit "yyy" value, and zero-based match - length is equal to (yyy + 8). - - If the second selection bit is equal to 1, the decoder uses binary tree - HighCoder to decode 8-bit "zzzzzzzz" value, and zero-based - match length is equal to (zzzzzzzz + 16). - -LZMA uses "posState" value as context to select the binary tree -from LowCoder and MidCoder binary tree arrays: - - unsigned posState = OutWindow.TotalPos & ((1 << pb) - 1); - -The full code of the length decoder: - -class CLenDecoder -{ - CProb Choice; - CProb Choice2; - CBitTreeDecoder<3> LowCoder[1 << kNumPosBitsMax]; - CBitTreeDecoder<3> MidCoder[1 << kNumPosBitsMax]; - CBitTreeDecoder<8> HighCoder; - -public: - - void Init() - { - Choice = PROB_INIT_VAL; - Choice2 = PROB_INIT_VAL; - HighCoder.Init(); - for (unsigned i = 0; i < (1 << kNumPosBitsMax); i++) - { - LowCoder[i].Init(); - MidCoder[i].Init(); - } - } - - unsigned Decode(CRangeDecoder *rc, unsigned posState) - { - if (rc->DecodeBit(&Choice) == 0) - return LowCoder[posState].Decode(rc); - if (rc->DecodeBit(&Choice2) == 0) - return 8 + MidCoder[posState].Decode(rc); - return 16 + HighCoder.Decode(rc); - } -}; - -The LZMA decoder uses two instances of CLenDecoder class. -The first instance is for the matches of "Simple Match" type, -and the second instance is for the matches of "Rep Match" type: - - CLenDecoder LenDecoder; - CLenDecoder RepLenDecoder; - - -The match distance decoding ---------------------------- - -LZMA supports dictionary sizes up to 4 GiB minus 1. -The value of match distance (decoded by distance decoder) can be -from 1 to 2^32. But the distance value that is equal to 2^32 is used to -indicate the "End of stream" marker. So real largest match distance -that is used for LZ-window match is (2^32 - 1). - -LZMA uses normalized match length (zero-based length) -to calculate the context state "lenState" do decode the distance value: - -#define kNumLenToPosStates 4 - - unsigned lenState = len; - if (lenState > kNumLenToPosStates - 1) - lenState = kNumLenToPosStates - 1; - -The distance decoder returns the "dist" value that is zero-based value -of match distance. The real match distance can be calculated with the -following code: - - matchDistance = dist + 1; - -The state of the distance decoder and the initialization code: - - #define kEndPosModelIndex 14 - #define kNumFullDistances (1 << (kEndPosModelIndex >> 1)) - #define kNumAlignBits 4 - - CBitTreeDecoder<6> PosSlotDecoder[kNumLenToPosStates]; - CProb PosDecoders[1 + kNumFullDistances - kEndPosModelIndex]; - CBitTreeDecoder AlignDecoder; - - void InitDist() - { - for (unsigned i = 0; i < kNumLenToPosStates; i++) - PosSlotDecoder[i].Init(); - AlignDecoder.Init(); - INIT_PROBS(PosDecoders); - } - -At first stage the distance decoder decodes 6-bit "posSlot" value with bit -tree decoder from PosSlotDecoder array. It's possible to get 2^6=64 different -"posSlot" values. - - unsigned posSlot = PosSlotDecoder[lenState].Decode(&RangeDec); - -The encoding scheme for distance value is shown in the following table: - -posSlot (decimal) / - zero-based distance (binary) - 0 0 - 1 1 - 2 10 - 3 11 - - 4 10 x - 5 11 x - 6 10 xx - 7 11 xx - 8 10 xxx - 9 11 xxx -10 10 xxxx -11 11 xxxx -12 10 xxxxx -13 11 xxxxx - -14 10 yy zzzz -15 11 yy zzzz -16 10 yyy zzzz -17 11 yyy zzzz -... -62 10 yyyyyyyyyyyyyyyyyyyyyyyyyy zzzz -63 11 yyyyyyyyyyyyyyyyyyyyyyyyyy zzzz - -where - "x ... x" means the sequence of binary symbols encoded with binary tree and - "Reverse" scheme. It uses separated binary tree for each posSlot from 4 to 13. - "y" means direct bit encoded with range coder. - "zzzz" means the sequence of four binary symbols encoded with binary - tree with "Reverse" scheme, where one common binary tree "AlignDecoder" - is used for all posSlot values. - -If (posSlot < 4), the "dist" value is equal to posSlot value. - -If (posSlot >= 4), the decoder uses "posSlot" value to calculate the value of - the high bits of "dist" value and the number of the low bits. - - If (4 <= posSlot < kEndPosModelIndex), the decoder uses bit tree decoders. - (one separated bit tree decoder per one posSlot value) and "Reverse" scheme. - In this implementation we use one CProb array "PosDecoders" that contains - all CProb variables for all these bit decoders. - - if (posSlot >= kEndPosModelIndex), the middle bits are decoded as direct - bits from RangeDecoder and the low 4 bits are decoded with a bit tree - decoder "AlignDecoder" with "Reverse" scheme. - -The code to decode zero-based match distance: - - unsigned DecodeDistance(unsigned len) - { - unsigned lenState = len; - if (lenState > kNumLenToPosStates - 1) - lenState = kNumLenToPosStates - 1; - - unsigned posSlot = PosSlotDecoder[lenState].Decode(&RangeDec); - if (posSlot < 4) - return posSlot; - - unsigned numDirectBits = (unsigned)((posSlot >> 1) - 1); - UInt32 dist = ((2 | (posSlot & 1)) << numDirectBits); - if (posSlot < kEndPosModelIndex) - dist += BitTreeReverseDecode(PosDecoders + dist - posSlot, numDirectBits, &RangeDec); - else - { - dist += RangeDec.DecodeDirectBits(numDirectBits - kNumAlignBits) << kNumAlignBits; - dist += AlignDecoder.ReverseDecode(&RangeDec); - } - return dist; - } - - - -LZMA Decoding modes -------------------- - -There are 2 types of LZMA streams: - -1) The stream with "End of stream" marker. -2) The stream without "End of stream" marker. - -And the LZMA Decoder supports 3 modes of decoding: - -1) The unpack size is undefined. The LZMA decoder stops decoding after - getting "End of stream" marker. - The input variables for that case: - - markerIsMandatory = true - unpackSizeDefined = false - unpackSize contains any value - -2) The unpack size is defined and LZMA decoder supports both variants, - where the stream can contain "End of stream" marker or the stream is - finished without "End of stream" marker. The LZMA decoder must detect - any of these situations. - The input variables for that case: - - markerIsMandatory = false - unpackSizeDefined = true - unpackSize contains unpack size - -3) The unpack size is defined and the LZMA stream must contain - "End of stream" marker - The input variables for that case: - - markerIsMandatory = true - unpackSizeDefined = true - unpackSize contains unpack size - - -The main loop of decoder ------------------------- - -The main loop of LZMA decoder: - -Initialize the LZMA state. -loop -{ - // begin of loop - Check "end of stream" conditions. - Decode Type of MATCH / LITERAL. - If it's LITERAL, decode LITERAL value and put the LITERAL to Window. - If it's MATCH, decode the length of match and the match distance. - Check error conditions, check end of stream conditions and copy - the sequence of match bytes from sliding window to current position - in window. - Go to begin of loop -} - -The reference implementation of LZMA decoder uses "unpackSize" variable -to keep the number of remaining bytes in output stream. So it reduces -"unpackSize" value after each decoded LITERAL or MATCH. - -The following code contains the "end of stream" condition check at the start -of the loop: - - if (unpackSizeDefined && unpackSize == 0 && !markerIsMandatory) - if (RangeDec.IsFinishedOK()) - return LZMA_RES_FINISHED_WITHOUT_MARKER; - -LZMA uses three types of matches: - -1) "Simple Match" - the match with distance value encoded with bit models. - -2) "Rep Match" - the match that uses the distance from distance - history table. - -3) "Short Rep Match" - the match of single byte length, that uses the latest - distance from distance history table. - -The LZMA decoder keeps the history of latest 4 match distances that were used -by decoder. That set of 4 variables contains zero-based match distances and -these variables are initialized with zero values: - - UInt32 rep0 = 0, rep1 = 0, rep2 = 0, rep3 = 0; - -The LZMA decoder uses binary model variables to select type of MATCH or LITERAL: - -#define kNumStates 12 -#define kNumPosBitsMax 4 - - CProb IsMatch[kNumStates << kNumPosBitsMax]; - CProb IsRep[kNumStates]; - CProb IsRepG0[kNumStates]; - CProb IsRepG1[kNumStates]; - CProb IsRepG2[kNumStates]; - CProb IsRep0Long[kNumStates << kNumPosBitsMax]; - -The decoder uses "state" variable value to select exact variable -from "IsRep", "IsRepG0", "IsRepG1" and "IsRepG2" arrays. -The "state" variable can get the value from 0 to 11. -Initial value for "state" variable is zero: - - unsigned state = 0; - -The "state" variable is updated after each LITERAL or MATCH with one of the -following functions: - -unsigned UpdateState_Literal(unsigned state) -{ - if (state < 4) return 0; - else if (state < 10) return state - 3; - else return state - 6; -} -unsigned UpdateState_Match (unsigned state) { return state < 7 ? 7 : 10; } -unsigned UpdateState_Rep (unsigned state) { return state < 7 ? 8 : 11; } -unsigned UpdateState_ShortRep(unsigned state) { return state < 7 ? 9 : 11; } - -The decoder calculates "state2" variable value to select exact variable from -"IsMatch" and "IsRep0Long" arrays: - -unsigned posState = OutWindow.TotalPos & ((1 << pb) - 1); -unsigned state2 = (state << kNumPosBitsMax) + posState; - -The decoder uses the following code flow scheme to select exact -type of LITERAL or MATCH: - -IsMatch[state2] decode - 0 - the Literal - 1 - the Match - IsRep[state] decode - 0 - Simple Match - 1 - Rep Match - IsRepG0[state] decode - 0 - the distance is rep0 - IsRep0Long[state2] decode - 0 - Short Rep Match - 1 - Rep Match 0 - 1 - - IsRepG1[state] decode - 0 - Rep Match 1 - 1 - - IsRepG2[state] decode - 0 - Rep Match 2 - 1 - Rep Match 3 - - -LITERAL symbol --------------- -If the value "0" was decoded with IsMatch[state2] decoding, we have "LITERAL" type. - -At first the LZMA decoder must check that it doesn't exceed -specified uncompressed size: - - if (unpackSizeDefined && unpackSize == 0) - return LZMA_RES_ERROR; - -Then it decodes literal value and puts it to sliding window: - - DecodeLiteral(state, rep0); - -Then the decoder must update the "state" value and "unpackSize" value; - - state = UpdateState_Literal(state); - unpackSize--; - -Then the decoder must go to the begin of main loop to decode next Match or Literal. - - -Simple Match ------------- - -If the value "1" was decoded with IsMatch[state2] decoding, -we have the "Simple Match" type. - -The distance history table is updated with the following scheme: - - rep3 = rep2; - rep2 = rep1; - rep1 = rep0; - -The zero-based length is decoded with "LenDecoder": - - len = LenDecoder.Decode(&RangeDec, posState); - -The state is update with UpdateState_Match function: - - state = UpdateState_Match(state); - -and the new "rep0" value is decoded with DecodeDistance: - - rep0 = DecodeDistance(len); - -That "rep0" will be used as zero-based distance for current match. - -If the value of "rep0" is equal to 0xFFFFFFFF, it means that we have -"End of stream" marker, so we can stop decoding and check finishing -condition in Range Decoder: - - if (rep0 == 0xFFFFFFFF) - return RangeDec.IsFinishedOK() ? - LZMA_RES_FINISHED_WITH_MARKER : - LZMA_RES_ERROR; - -If uncompressed size is defined, LZMA decoder must check that it doesn't -exceed that specified uncompressed size: - - if (unpackSizeDefined && unpackSize == 0) - return LZMA_RES_ERROR; - -Also the decoder must check that "rep0" value is not larger than dictionary size -and is not larger than the number of already decoded bytes: - - if (rep0 >= dictSize || !OutWindow.CheckDistance(rep0)) - return LZMA_RES_ERROR; - -Then the decoder must copy match bytes as described in -"The match symbols copying" section. - - -Rep Match ---------- - -If the LZMA decoder has decoded the value "1" with IsRep[state] variable, -we have "Rep Match" type. - -At first the LZMA decoder must check that it doesn't exceed -specified uncompressed size: - - if (unpackSizeDefined && unpackSize == 0) - return LZMA_RES_ERROR; - -Also the decoder must return error, if the LZ window is empty: - - if (OutWindow.IsEmpty()) - return LZMA_RES_ERROR; - -If the match type is "Rep Match", the decoder uses one of the 4 variables of -distance history table to get the value of distance for current match. -And there are 4 corresponding ways of decoding flow. - -The decoder updates the distance history with the following scheme -depending from type of match: - -- "Rep Match 0" or "Short Rep Match": - ; LZMA doesn't update the distance history - -- "Rep Match 1": - UInt32 dist = rep1; - rep1 = rep0; - rep0 = dist; - -- "Rep Match 2": - UInt32 dist = rep2; - rep2 = rep1; - rep1 = rep0; - rep0 = dist; - -- "Rep Match 3": - UInt32 dist = rep3; - rep3 = rep2; - rep2 = rep1; - rep1 = rep0; - rep0 = dist; - -Then the decoder decodes exact subtype of "Rep Match" using "IsRepG0", "IsRep0Long", -"IsRepG1", "IsRepG2". - -If the subtype is "Short Rep Match", the decoder updates the state, puts -the one byte from window to current position in window and goes to next -MATCH/LITERAL symbol (the begin of main loop): - - state = UpdateState_ShortRep(state); - OutWindow.PutByte(OutWindow.GetByte(rep0 + 1)); - unpackSize--; - continue; - -In other cases (Rep Match 0/1/2/3), it decodes the zero-based -length of match with "RepLenDecoder" decoder: - - len = RepLenDecoder.Decode(&RangeDec, posState); - -Then it updates the state: - - state = UpdateState_Rep(state); - -Then the decoder must copy match bytes as described in -"The Match symbols copying" section. - - -The match symbols copying -------------------------- - -If we have the match (Simple Match or Rep Match 0/1/2/3), the decoder must -copy the sequence of bytes with calculated match distance and match length. -If uncompressed size is defined, LZMA decoder must check that it doesn't -exceed that specified uncompressed size: - - len += kMatchMinLen; - bool isError = false; - if (unpackSizeDefined && unpackSize < len) - { - len = (unsigned)unpackSize; - isError = true; - } - OutWindow.CopyMatch(rep0 + 1, len); - unpackSize -= len; - if (isError) - return LZMA_RES_ERROR; - -Then the decoder must go to the begin of main loop to decode next MATCH or LITERAL. - - - -NOTES ------ - -This specification doesn't describe the variant of decoder implementation -that supports partial decoding. Such partial decoding case can require some -changes in "end of stream" condition checks code. Also such code -can use additional status codes, returned by decoder. - -This specification uses C++ code with templates to simplify describing. -The optimized version of LZMA decoder doesn't need templates. -Such optimized version can use just two arrays of CProb variables: - 1) The dynamic array of CProb variables allocated for the Literal Decoder. - 2) The one common array that contains all other CProb variables. - - -References: - -1. G. N. N. Martin, Range encoding: an algorithm for removing redundancy - from a digitized message, Video & Data Recording Conference, - Southampton, UK, July 24-27, 1979. +LZMA specification (DRAFT version) +---------------------------------- + +Author: Igor Pavlov +Date: 2015-06-14 + +This specification defines the format of LZMA compressed data and lzma file format. + +Notation +-------- + +We use the syntax of C++ programming language. +We use the following types in C++ code: + unsigned - unsigned integer, at least 16 bits in size + int - signed integer, at least 16 bits in size + UInt64 - 64-bit unsigned integer + UInt32 - 32-bit unsigned integer + UInt16 - 16-bit unsigned integer + Byte - 8-bit unsigned integer + bool - boolean type with two possible values: false, true + + +lzma file format +================ + +The lzma file contains the raw LZMA stream and the header with related properties. + +The files in that format use ".lzma" extension. + +The lzma file format layout: + +Offset Size Description + + 0 1 LZMA model properties (lc, lp, pb) in encoded form + 1 4 Dictionary size (32-bit unsigned integer, little-endian) + 5 8 Uncompressed size (64-bit unsigned integer, little-endian) + 13 Compressed data (LZMA stream) + +LZMA properties: + + name Range Description + + lc [0, 8] the number of "literal context" bits + lp [0, 4] the number of "literal pos" bits + pb [0, 4] the number of "pos" bits +dictSize [0, 2^32 - 1] the dictionary size + +The following code encodes LZMA properties: + +void EncodeProperties(Byte *properties) +{ + properties[0] = (Byte)((pb * 5 + lp) * 9 + lc); + Set_UInt32_LittleEndian(properties + 1, dictSize); +} + +If the value of dictionary size in properties is smaller than (1 << 12), +the LZMA decoder must set the dictionary size variable to (1 << 12). + +#define LZMA_DIC_MIN (1 << 12) + + unsigned lc, pb, lp; + UInt32 dictSize; + UInt32 dictSizeInProperties; + + void DecodeProperties(const Byte *properties) + { + unsigned d = properties[0]; + if (d >= (9 * 5 * 5)) + throw "Incorrect LZMA properties"; + lc = d % 9; + d /= 9; + pb = d / 5; + lp = d % 5; + dictSizeInProperties = 0; + for (int i = 0; i < 4; i++) + dictSizeInProperties |= (UInt32)properties[i + 1] << (8 * i); + dictSize = dictSizeInProperties; + if (dictSize < LZMA_DIC_MIN) + dictSize = LZMA_DIC_MIN; + } + +If "Uncompressed size" field contains ones in all 64 bits, it means that +uncompressed size is unknown and there is the "end marker" in stream, +that indicates the end of decoding point. +In opposite case, if the value from "Uncompressed size" field is not +equal to ((2^64) - 1), the LZMA stream decoding must be finished after +specified number of bytes (Uncompressed size) is decoded. And if there +is the "end marker", the LZMA decoder must read that marker also. + + +The new scheme to encode LZMA properties +---------------------------------------- + +If LZMA compression is used for some another format, it's recommended to +use a new improved scheme to encode LZMA properties. That new scheme was +used in xz format that uses the LZMA2 compression algorithm. +The LZMA2 is a new compression algorithm that is based on the LZMA algorithm. + +The dictionary size in LZMA2 is encoded with just one byte and LZMA2 supports +only reduced set of dictionary sizes: + (2 << 11), (3 << 11), + (2 << 12), (3 << 12), + ... + (2 << 30), (3 << 30), + (2 << 31) - 1 + +The dictionary size can be extracted from encoded value with the following code: + + dictSize = (p == 40) ? 0xFFFFFFFF : (((UInt32)2 | ((p) & 1)) << ((p) / 2 + 11)); + +Also there is additional limitation (lc + lp <= 4) in LZMA2 for values of +"lc" and "lp" properties: + + if (lc + lp > 4) + throw "Unsupported properties: (lc + lp) > 4"; + +There are some advantages for LZMA decoder with such (lc + lp) value +limitation. It reduces the maximum size of tables allocated by decoder. +And it reduces the complexity of initialization procedure, that can be +important to keep high speed of decoding of big number of small LZMA streams. + +It's recommended to use that limitation (lc + lp <= 4) for any new format +that uses LZMA compression. Note that the combinations of "lc" and "lp" +parameters, where (lc + lp > 4), can provide significant improvement in +compression ratio only in some rare cases. + +The LZMA properties can be encoded into two bytes in new scheme: + +Offset Size Description + + 0 1 The dictionary size encoded with LZMA2 scheme + 1 1 LZMA model properties (lc, lp, pb) in encoded form + + +The RAM usage +============= + +The RAM usage for LZMA decoder is determined by the following parts: + +1) The Sliding Window (from 4 KiB to 4 GiB). +2) The probability model counter arrays (arrays of 16-bit variables). +3) Some additional state variables (about 10 variables of 32-bit integers). + + +The RAM usage for Sliding Window +-------------------------------- + +There are two main scenarios of decoding: + +1) The decoding of full stream to one RAM buffer. + + If we decode full LZMA stream to one output buffer in RAM, the decoder + can use that output buffer as sliding window. So the decoder doesn't + need additional buffer allocated for sliding window. + +2) The decoding to some external storage. + + If we decode LZMA stream to external storage, the decoder must allocate + the buffer for sliding window. The size of that buffer must be equal + or larger than the value of dictionary size from properties of LZMA stream. + +In this specification we describe the code for decoding to some external +storage. The optimized version of code for decoding of full stream to one +output RAM buffer can require some minor changes in code. + + +The RAM usage for the probability model counters +------------------------------------------------ + +The size of the probability model counter arrays is calculated with the +following formula: + +size_of_prob_arrays = 1846 + 768 * (1 << (lp + lc)) + +Each probability model counter is 11-bit unsigned integer. +If we use 16-bit integer variables (2-byte integers) for these probability +model counters, the RAM usage required by probability model counter arrays +can be estimated with the following formula: + + RAM = 4 KiB + 1.5 KiB * (1 << (lp + lc)) + +For example, for default LZMA parameters (lp = 0 and lc = 3), the RAM usage is + + RAM_lc3_lp0 = 4 KiB + 1.5 KiB * 8 = 16 KiB + +The maximum RAM state usage is required for decoding the stream with lp = 4 +and lc = 8: + + RAM_lc8_lp4 = 4 KiB + 1.5 KiB * 4096 = 6148 KiB + +If the decoder uses LZMA2's limited property condition +(lc + lp <= 4), the RAM usage will be not larger than + + RAM_lc_lp_4 = 4 KiB + 1.5 KiB * 16 = 28 KiB + + +The RAM usage for encoder +------------------------- + +There are many variants for LZMA encoding code. +These variants have different values for memory consumption. +Note that memory consumption for LZMA Encoder can not be +smaller than memory consumption of LZMA Decoder for same stream. + +The RAM usage required by modern effective implementation of +LZMA Encoder can be estimated with the following formula: + + Encoder_RAM_Usage = 4 MiB + 11 * dictionarySize. + +But there are some modes of the encoder that require less memory. + + +LZMA Decoding +============= + +The LZMA compression algorithm uses LZ-based compression with Sliding Window +and Range Encoding as entropy coding method. + + +Sliding Window +-------------- + +LZMA uses Sliding Window compression similar to LZ77 algorithm. + +LZMA stream must be decoded to the sequence that consists +of MATCHES and LITERALS: + + - a LITERAL is a 8-bit character (one byte). + The decoder just puts that LITERAL to the uncompressed stream. + + - a MATCH is a pair of two numbers (DISTANCE-LENGTH pair). + The decoder takes one byte exactly "DISTANCE" characters behind + current position in the uncompressed stream and puts it to + uncompressed stream. The decoder must repeat it "LENGTH" times. + +The "DISTANCE" can not be larger than dictionary size. +And the "DISTANCE" can not be larger than the number of bytes in +the uncompressed stream that were decoded before that match. + +In this specification we use cyclic buffer to implement Sliding Window +for LZMA decoder: + +class COutWindow +{ + Byte *Buf; + UInt32 Pos; + UInt32 Size; + bool IsFull; + +public: + unsigned TotalPos; + COutStream OutStream; + + COutWindow(): Buf(NULL) {} + ~COutWindow() { delete []Buf; } + + void Create(UInt32 dictSize) + { + Buf = new Byte[dictSize]; + Pos = 0; + Size = dictSize; + IsFull = false; + TotalPos = 0; + } + + void PutByte(Byte b) + { + TotalPos++; + Buf[Pos++] = b; + if (Pos == Size) + { + Pos = 0; + IsFull = true; + } + OutStream.WriteByte(b); + } + + Byte GetByte(UInt32 dist) const + { + return Buf[dist <= Pos ? Pos - dist : Size - dist + Pos]; + } + + void CopyMatch(UInt32 dist, unsigned len) + { + for (; len > 0; len--) + PutByte(GetByte(dist)); + } + + bool CheckDistance(UInt32 dist) const + { + return dist <= Pos || IsFull; + } + + bool IsEmpty() const + { + return Pos == 0 && !IsFull; + } +}; + + +In another implementation it's possible to use one buffer that contains +Sliding Window and the whole data stream after uncompressing. + + +Range Decoder +------------- + +LZMA algorithm uses Range Encoding (1) as entropy coding method. + +LZMA stream contains just one very big number in big-endian encoding. +LZMA decoder uses the Range Decoder to extract a sequence of binary +symbols from that big number. + +The state of the Range Decoder: + +struct CRangeDecoder +{ + UInt32 Range; + UInt32 Code; + InputStream *InStream; + + bool Corrupted; +} + +The notes about UInt32 type for the "Range" and "Code" variables: + + It's possible to use 64-bit (unsigned or signed) integer type + for the "Range" and the "Code" variables instead of 32-bit unsigned, + but some additional code must be used to truncate the values to + low 32-bits after some operations. + + If the programming language does not support 32-bit unsigned integer type + (like in case of JAVA language), it's possible to use 32-bit signed integer, + but some code must be changed. For example, it's required to change the code + that uses comparison operations for UInt32 variables in this specification. + +The Range Decoder can be in some states that can be treated as +"Corruption" in LZMA stream. The Range Decoder uses the variable "Corrupted": + + (Corrupted == false), if the Range Decoder has not detected any corruption. + (Corrupted == true), if the Range Decoder has detected some corruption. + +The reference LZMA Decoder ignores the value of the "Corrupted" variable. +So it continues to decode the stream, even if the corruption can be detected +in the Range Decoder. To provide the full compatibility with output of the +reference LZMA Decoder, another LZMA Decoder implementations must also +ignore the value of the "Corrupted" variable. + +The LZMA Encoder is required to create only such LZMA streams, that will not +lead the Range Decoder to states, where the "Corrupted" variable is set to true. + +The Range Decoder reads first 5 bytes from input stream to initialize +the state: + +bool CRangeDecoder::Init() +{ + Corrupted = false; + Range = 0xFFFFFFFF; + Code = 0; + + Byte b = InStream->ReadByte(); + + for (int i = 0; i < 4; i++) + Code = (Code << 8) | InStream->ReadByte(); + + if (b != 0 || Code == Range) + Corrupted = true; + return b == 0; +} + +The LZMA Encoder always writes ZERO in initial byte of compressed stream. +That scheme allows to simplify the code of the Range Encoder in the +LZMA Encoder. If initial byte is not equal to ZERO, the LZMA Decoder must +stop decoding and report error. + +After the last bit of data was decoded by Range Decoder, the value of the +"Code" variable must be equal to 0. The LZMA Decoder must check it by +calling the IsFinishedOK() function: + + bool IsFinishedOK() const { return Code == 0; } + +If there is corruption in data stream, there is big probability that +the "Code" value will be not equal to 0 in the Finish() function. So that +check in the IsFinishedOK() function provides very good feature for +corruption detection. + +The value of the "Range" variable before each bit decoding can not be smaller +than ((UInt32)1 << 24). The Normalize() function keeps the "Range" value in +described range. + +#define kTopValue ((UInt32)1 << 24) + +void CRangeDecoder::Normalize() +{ + if (Range < kTopValue) + { + Range <<= 8; + Code = (Code << 8) | InStream->ReadByte(); + } +} + +Notes: if the size of the "Code" variable is larger than 32 bits, it's +required to keep only low 32 bits of the "Code" variable after the change +in Normalize() function. + +If the LZMA Stream is not corrupted, the value of the "Code" variable is +always smaller than value of the "Range" variable. +But the Range Decoder ignores some types of corruptions, so the value of +the "Code" variable can be equal or larger than value of the "Range" variable +for some "Corrupted" archives. + + +LZMA uses Range Encoding only with binary symbols of two types: + 1) binary symbols with fixed and equal probabilities (direct bits) + 2) binary symbols with predicted probabilities + +The DecodeDirectBits() function decodes the sequence of direct bits: + +UInt32 CRangeDecoder::DecodeDirectBits(unsigned numBits) +{ + UInt32 res = 0; + do + { + Range >>= 1; + Code -= Range; + UInt32 t = 0 - ((UInt32)Code >> 31); + Code += Range & t; + + if (Code == Range) + Corrupted = true; + + Normalize(); + res <<= 1; + res += t + 1; + } + while (--numBits); + return res; +} + + +The Bit Decoding with Probability Model +--------------------------------------- + +The task of Bit Probability Model is to estimate probabilities of binary +symbols. And then it provides the Range Decoder with that information. +The better prediction provides better compression ratio. +The Bit Probability Model uses statistical data of previous decoded +symbols. + +That estimated probability is presented as 11-bit unsigned integer value +that represents the probability of symbol "0". + +#define kNumBitModelTotalBits 11 + +Mathematical probabilities can be presented with the following formulas: + probability(symbol_0) = prob / 2048. + probability(symbol_1) = 1 - Probability(symbol_0) = + = 1 - prob / 2048 = + = (2048 - prob) / 2048 +where the "prob" variable contains 11-bit integer probability counter. + +It's recommended to use 16-bit unsigned integer type, to store these 11-bit +probability values: + +typedef UInt16 CProb; + +Each probability value must be initialized with value ((1 << 11) / 2), +that represents the state, where probabilities of symbols 0 and 1 +are equal to 0.5: + +#define PROB_INIT_VAL ((1 << kNumBitModelTotalBits) / 2) + +The INIT_PROBS macro is used to initialize the array of CProb variables: + +#define INIT_PROBS(p) \ + { for (unsigned i = 0; i < sizeof(p) / sizeof(p[0]); i++) p[i] = PROB_INIT_VAL; } + + +The DecodeBit() function decodes one bit. +The LZMA decoder provides the pointer to CProb variable that contains +information about estimated probability for symbol 0 and the Range Decoder +updates that CProb variable after decoding. The Range Decoder increases +estimated probability of the symbol that was decoded: + +#define kNumMoveBits 5 + +unsigned CRangeDecoder::DecodeBit(CProb *prob) +{ + unsigned v = *prob; + UInt32 bound = (Range >> kNumBitModelTotalBits) * v; + unsigned symbol; + if (Code < bound) + { + v += ((1 << kNumBitModelTotalBits) - v) >> kNumMoveBits; + Range = bound; + symbol = 0; + } + else + { + v -= v >> kNumMoveBits; + Code -= bound; + Range -= bound; + symbol = 1; + } + *prob = (CProb)v; + Normalize(); + return symbol; +} + + +The Binary Tree of bit model counters +------------------------------------- + +LZMA uses a tree of Bit model variables to decode symbol that needs +several bits for storing. There are two versions of such trees in LZMA: + 1) the tree that decodes bits from high bit to low bit (the normal scheme). + 2) the tree that decodes bits from low bit to high bit (the reverse scheme). + +Each binary tree structure supports different size of decoded symbol +(the size of binary sequence that contains value of symbol). +If that size of decoded symbol is "NumBits" bits, the tree structure +uses the array of (2 << NumBits) counters of CProb type. +But only ((2 << NumBits) - 1) items are used by encoder and decoder. +The first item (the item with index equal to 0) in array is unused. +That scheme with unused array's item allows to simplify the code. + +unsigned BitTreeReverseDecode(CProb *probs, unsigned numBits, CRangeDecoder *rc) +{ + unsigned m = 1; + unsigned symbol = 0; + for (unsigned i = 0; i < numBits; i++) + { + unsigned bit = rc->DecodeBit(&probs[m]); + m <<= 1; + m += bit; + symbol |= (bit << i); + } + return symbol; +} + +template +class CBitTreeDecoder +{ + CProb Probs[(unsigned)1 << NumBits]; + +public: + + void Init() + { + INIT_PROBS(Probs); + } + + unsigned Decode(CRangeDecoder *rc) + { + unsigned m = 1; + for (unsigned i = 0; i < NumBits; i++) + m = (m << 1) + rc->DecodeBit(&Probs[m]); + return m - ((unsigned)1 << NumBits); + } + + unsigned ReverseDecode(CRangeDecoder *rc) + { + return BitTreeReverseDecode(Probs, NumBits, rc); + } +}; + + +LZ part of LZMA +--------------- + +LZ part of LZMA describes details about the decoding of MATCHES and LITERALS. + + +The Literal Decoding +-------------------- + +The LZMA Decoder uses (1 << (lc + lp)) tables with CProb values, where +each table contains 0x300 CProb values: + + CProb *LitProbs; + + void CreateLiterals() + { + LitProbs = new CProb[(UInt32)0x300 << (lc + lp)]; + } + + void InitLiterals() + { + UInt32 num = (UInt32)0x300 << (lc + lp); + for (UInt32 i = 0; i < num; i++) + LitProbs[i] = PROB_INIT_VAL; + } + +To select the table for decoding it uses the context that consists of +(lc) high bits from previous literal and (lp) low bits from value that +represents current position in outputStream. + +If (State > 7), the Literal Decoder also uses "matchByte" that represents +the byte in OutputStream at position the is the DISTANCE bytes before +current position, where the DISTANCE is the distance in DISTANCE-LENGTH pair +of latest decoded match. + +The following code decodes one literal and puts it to Sliding Window buffer: + + void DecodeLiteral(unsigned state, UInt32 rep0) + { + unsigned prevByte = 0; + if (!OutWindow.IsEmpty()) + prevByte = OutWindow.GetByte(1); + + unsigned symbol = 1; + unsigned litState = ((OutWindow.TotalPos & ((1 << lp) - 1)) << lc) + (prevByte >> (8 - lc)); + CProb *probs = &LitProbs[(UInt32)0x300 * litState]; + + if (state >= 7) + { + unsigned matchByte = OutWindow.GetByte(rep0 + 1); + do + { + unsigned matchBit = (matchByte >> 7) & 1; + matchByte <<= 1; + unsigned bit = RangeDec.DecodeBit(&probs[((1 + matchBit) << 8) + symbol]); + symbol = (symbol << 1) | bit; + if (matchBit != bit) + break; + } + while (symbol < 0x100); + } + while (symbol < 0x100) + symbol = (symbol << 1) | RangeDec.DecodeBit(&probs[symbol]); + OutWindow.PutByte((Byte)(symbol - 0x100)); + } + + +The match length decoding +------------------------- + +The match length decoder returns normalized (zero-based value) +length of match. That value can be converted to real length of the match +with the following code: + +#define kMatchMinLen 2 + + matchLen = len + kMatchMinLen; + +The match length decoder can return the values from 0 to 271. +And the corresponded real match length values can be in the range +from 2 to 273. + +The following scheme is used for the match length encoding: + + Binary encoding Binary Tree structure Zero-based match length + sequence (binary + decimal): + + 0 xxx LowCoder[posState] xxx + 1 0 yyy MidCoder[posState] yyy + 8 + 1 1 zzzzzzzz HighCoder zzzzzzzz + 16 + +LZMA uses bit model variable "Choice" to decode the first selection bit. + +If the first selection bit is equal to 0, the decoder uses binary tree + LowCoder[posState] to decode 3-bit zero-based match length (xxx). + +If the first selection bit is equal to 1, the decoder uses bit model + variable "Choice2" to decode the second selection bit. + + If the second selection bit is equal to 0, the decoder uses binary tree + MidCoder[posState] to decode 3-bit "yyy" value, and zero-based match + length is equal to (yyy + 8). + + If the second selection bit is equal to 1, the decoder uses binary tree + HighCoder to decode 8-bit "zzzzzzzz" value, and zero-based + match length is equal to (zzzzzzzz + 16). + +LZMA uses "posState" value as context to select the binary tree +from LowCoder and MidCoder binary tree arrays: + + unsigned posState = OutWindow.TotalPos & ((1 << pb) - 1); + +The full code of the length decoder: + +class CLenDecoder +{ + CProb Choice; + CProb Choice2; + CBitTreeDecoder<3> LowCoder[1 << kNumPosBitsMax]; + CBitTreeDecoder<3> MidCoder[1 << kNumPosBitsMax]; + CBitTreeDecoder<8> HighCoder; + +public: + + void Init() + { + Choice = PROB_INIT_VAL; + Choice2 = PROB_INIT_VAL; + HighCoder.Init(); + for (unsigned i = 0; i < (1 << kNumPosBitsMax); i++) + { + LowCoder[i].Init(); + MidCoder[i].Init(); + } + } + + unsigned Decode(CRangeDecoder *rc, unsigned posState) + { + if (rc->DecodeBit(&Choice) == 0) + return LowCoder[posState].Decode(rc); + if (rc->DecodeBit(&Choice2) == 0) + return 8 + MidCoder[posState].Decode(rc); + return 16 + HighCoder.Decode(rc); + } +}; + +The LZMA decoder uses two instances of CLenDecoder class. +The first instance is for the matches of "Simple Match" type, +and the second instance is for the matches of "Rep Match" type: + + CLenDecoder LenDecoder; + CLenDecoder RepLenDecoder; + + +The match distance decoding +--------------------------- + +LZMA supports dictionary sizes up to 4 GiB minus 1. +The value of match distance (decoded by distance decoder) can be +from 1 to 2^32. But the distance value that is equal to 2^32 is used to +indicate the "End of stream" marker. So real largest match distance +that is used for LZ-window match is (2^32 - 1). + +LZMA uses normalized match length (zero-based length) +to calculate the context state "lenState" do decode the distance value: + +#define kNumLenToPosStates 4 + + unsigned lenState = len; + if (lenState > kNumLenToPosStates - 1) + lenState = kNumLenToPosStates - 1; + +The distance decoder returns the "dist" value that is zero-based value +of match distance. The real match distance can be calculated with the +following code: + + matchDistance = dist + 1; + +The state of the distance decoder and the initialization code: + + #define kEndPosModelIndex 14 + #define kNumFullDistances (1 << (kEndPosModelIndex >> 1)) + #define kNumAlignBits 4 + + CBitTreeDecoder<6> PosSlotDecoder[kNumLenToPosStates]; + CProb PosDecoders[1 + kNumFullDistances - kEndPosModelIndex]; + CBitTreeDecoder AlignDecoder; + + void InitDist() + { + for (unsigned i = 0; i < kNumLenToPosStates; i++) + PosSlotDecoder[i].Init(); + AlignDecoder.Init(); + INIT_PROBS(PosDecoders); + } + +At first stage the distance decoder decodes 6-bit "posSlot" value with bit +tree decoder from PosSlotDecoder array. It's possible to get 2^6=64 different +"posSlot" values. + + unsigned posSlot = PosSlotDecoder[lenState].Decode(&RangeDec); + +The encoding scheme for distance value is shown in the following table: + +posSlot (decimal) / + zero-based distance (binary) + 0 0 + 1 1 + 2 10 + 3 11 + + 4 10 x + 5 11 x + 6 10 xx + 7 11 xx + 8 10 xxx + 9 11 xxx +10 10 xxxx +11 11 xxxx +12 10 xxxxx +13 11 xxxxx + +14 10 yy zzzz +15 11 yy zzzz +16 10 yyy zzzz +17 11 yyy zzzz +... +62 10 yyyyyyyyyyyyyyyyyyyyyyyyyy zzzz +63 11 yyyyyyyyyyyyyyyyyyyyyyyyyy zzzz + +where + "x ... x" means the sequence of binary symbols encoded with binary tree and + "Reverse" scheme. It uses separated binary tree for each posSlot from 4 to 13. + "y" means direct bit encoded with range coder. + "zzzz" means the sequence of four binary symbols encoded with binary + tree with "Reverse" scheme, where one common binary tree "AlignDecoder" + is used for all posSlot values. + +If (posSlot < 4), the "dist" value is equal to posSlot value. + +If (posSlot >= 4), the decoder uses "posSlot" value to calculate the value of + the high bits of "dist" value and the number of the low bits. + + If (4 <= posSlot < kEndPosModelIndex), the decoder uses bit tree decoders. + (one separated bit tree decoder per one posSlot value) and "Reverse" scheme. + In this implementation we use one CProb array "PosDecoders" that contains + all CProb variables for all these bit decoders. + + if (posSlot >= kEndPosModelIndex), the middle bits are decoded as direct + bits from RangeDecoder and the low 4 bits are decoded with a bit tree + decoder "AlignDecoder" with "Reverse" scheme. + +The code to decode zero-based match distance: + + unsigned DecodeDistance(unsigned len) + { + unsigned lenState = len; + if (lenState > kNumLenToPosStates - 1) + lenState = kNumLenToPosStates - 1; + + unsigned posSlot = PosSlotDecoder[lenState].Decode(&RangeDec); + if (posSlot < 4) + return posSlot; + + unsigned numDirectBits = (unsigned)((posSlot >> 1) - 1); + UInt32 dist = ((2 | (posSlot & 1)) << numDirectBits); + if (posSlot < kEndPosModelIndex) + dist += BitTreeReverseDecode(PosDecoders + dist - posSlot, numDirectBits, &RangeDec); + else + { + dist += RangeDec.DecodeDirectBits(numDirectBits - kNumAlignBits) << kNumAlignBits; + dist += AlignDecoder.ReverseDecode(&RangeDec); + } + return dist; + } + + + +LZMA Decoding modes +------------------- + +There are 2 types of LZMA streams: + +1) The stream with "End of stream" marker. +2) The stream without "End of stream" marker. + +And the LZMA Decoder supports 3 modes of decoding: + +1) The unpack size is undefined. The LZMA decoder stops decoding after + getting "End of stream" marker. + The input variables for that case: + + markerIsMandatory = true + unpackSizeDefined = false + unpackSize contains any value + +2) The unpack size is defined and LZMA decoder supports both variants, + where the stream can contain "End of stream" marker or the stream is + finished without "End of stream" marker. The LZMA decoder must detect + any of these situations. + The input variables for that case: + + markerIsMandatory = false + unpackSizeDefined = true + unpackSize contains unpack size + +3) The unpack size is defined and the LZMA stream must contain + "End of stream" marker + The input variables for that case: + + markerIsMandatory = true + unpackSizeDefined = true + unpackSize contains unpack size + + +The main loop of decoder +------------------------ + +The main loop of LZMA decoder: + +Initialize the LZMA state. +loop +{ + // begin of loop + Check "end of stream" conditions. + Decode Type of MATCH / LITERAL. + If it's LITERAL, decode LITERAL value and put the LITERAL to Window. + If it's MATCH, decode the length of match and the match distance. + Check error conditions, check end of stream conditions and copy + the sequence of match bytes from sliding window to current position + in window. + Go to begin of loop +} + +The reference implementation of LZMA decoder uses "unpackSize" variable +to keep the number of remaining bytes in output stream. So it reduces +"unpackSize" value after each decoded LITERAL or MATCH. + +The following code contains the "end of stream" condition check at the start +of the loop: + + if (unpackSizeDefined && unpackSize == 0 && !markerIsMandatory) + if (RangeDec.IsFinishedOK()) + return LZMA_RES_FINISHED_WITHOUT_MARKER; + +LZMA uses three types of matches: + +1) "Simple Match" - the match with distance value encoded with bit models. + +2) "Rep Match" - the match that uses the distance from distance + history table. + +3) "Short Rep Match" - the match of single byte length, that uses the latest + distance from distance history table. + +The LZMA decoder keeps the history of latest 4 match distances that were used +by decoder. That set of 4 variables contains zero-based match distances and +these variables are initialized with zero values: + + UInt32 rep0 = 0, rep1 = 0, rep2 = 0, rep3 = 0; + +The LZMA decoder uses binary model variables to select type of MATCH or LITERAL: + +#define kNumStates 12 +#define kNumPosBitsMax 4 + + CProb IsMatch[kNumStates << kNumPosBitsMax]; + CProb IsRep[kNumStates]; + CProb IsRepG0[kNumStates]; + CProb IsRepG1[kNumStates]; + CProb IsRepG2[kNumStates]; + CProb IsRep0Long[kNumStates << kNumPosBitsMax]; + +The decoder uses "state" variable value to select exact variable +from "IsRep", "IsRepG0", "IsRepG1" and "IsRepG2" arrays. +The "state" variable can get the value from 0 to 11. +Initial value for "state" variable is zero: + + unsigned state = 0; + +The "state" variable is updated after each LITERAL or MATCH with one of the +following functions: + +unsigned UpdateState_Literal(unsigned state) +{ + if (state < 4) return 0; + else if (state < 10) return state - 3; + else return state - 6; +} +unsigned UpdateState_Match (unsigned state) { return state < 7 ? 7 : 10; } +unsigned UpdateState_Rep (unsigned state) { return state < 7 ? 8 : 11; } +unsigned UpdateState_ShortRep(unsigned state) { return state < 7 ? 9 : 11; } + +The decoder calculates "state2" variable value to select exact variable from +"IsMatch" and "IsRep0Long" arrays: + +unsigned posState = OutWindow.TotalPos & ((1 << pb) - 1); +unsigned state2 = (state << kNumPosBitsMax) + posState; + +The decoder uses the following code flow scheme to select exact +type of LITERAL or MATCH: + +IsMatch[state2] decode + 0 - the Literal + 1 - the Match + IsRep[state] decode + 0 - Simple Match + 1 - Rep Match + IsRepG0[state] decode + 0 - the distance is rep0 + IsRep0Long[state2] decode + 0 - Short Rep Match + 1 - Rep Match 0 + 1 - + IsRepG1[state] decode + 0 - Rep Match 1 + 1 - + IsRepG2[state] decode + 0 - Rep Match 2 + 1 - Rep Match 3 + + +LITERAL symbol +-------------- +If the value "0" was decoded with IsMatch[state2] decoding, we have "LITERAL" type. + +At first the LZMA decoder must check that it doesn't exceed +specified uncompressed size: + + if (unpackSizeDefined && unpackSize == 0) + return LZMA_RES_ERROR; + +Then it decodes literal value and puts it to sliding window: + + DecodeLiteral(state, rep0); + +Then the decoder must update the "state" value and "unpackSize" value; + + state = UpdateState_Literal(state); + unpackSize--; + +Then the decoder must go to the begin of main loop to decode next Match or Literal. + + +Simple Match +------------ + +If the value "1" was decoded with IsMatch[state2] decoding, +we have the "Simple Match" type. + +The distance history table is updated with the following scheme: + + rep3 = rep2; + rep2 = rep1; + rep1 = rep0; + +The zero-based length is decoded with "LenDecoder": + + len = LenDecoder.Decode(&RangeDec, posState); + +The state is update with UpdateState_Match function: + + state = UpdateState_Match(state); + +and the new "rep0" value is decoded with DecodeDistance: + + rep0 = DecodeDistance(len); + +That "rep0" will be used as zero-based distance for current match. + +If the value of "rep0" is equal to 0xFFFFFFFF, it means that we have +"End of stream" marker, so we can stop decoding and check finishing +condition in Range Decoder: + + if (rep0 == 0xFFFFFFFF) + return RangeDec.IsFinishedOK() ? + LZMA_RES_FINISHED_WITH_MARKER : + LZMA_RES_ERROR; + +If uncompressed size is defined, LZMA decoder must check that it doesn't +exceed that specified uncompressed size: + + if (unpackSizeDefined && unpackSize == 0) + return LZMA_RES_ERROR; + +Also the decoder must check that "rep0" value is not larger than dictionary size +and is not larger than the number of already decoded bytes: + + if (rep0 >= dictSize || !OutWindow.CheckDistance(rep0)) + return LZMA_RES_ERROR; + +Then the decoder must copy match bytes as described in +"The match symbols copying" section. + + +Rep Match +--------- + +If the LZMA decoder has decoded the value "1" with IsRep[state] variable, +we have "Rep Match" type. + +At first the LZMA decoder must check that it doesn't exceed +specified uncompressed size: + + if (unpackSizeDefined && unpackSize == 0) + return LZMA_RES_ERROR; + +Also the decoder must return error, if the LZ window is empty: + + if (OutWindow.IsEmpty()) + return LZMA_RES_ERROR; + +If the match type is "Rep Match", the decoder uses one of the 4 variables of +distance history table to get the value of distance for current match. +And there are 4 corresponding ways of decoding flow. + +The decoder updates the distance history with the following scheme +depending from type of match: + +- "Rep Match 0" or "Short Rep Match": + ; LZMA doesn't update the distance history + +- "Rep Match 1": + UInt32 dist = rep1; + rep1 = rep0; + rep0 = dist; + +- "Rep Match 2": + UInt32 dist = rep2; + rep2 = rep1; + rep1 = rep0; + rep0 = dist; + +- "Rep Match 3": + UInt32 dist = rep3; + rep3 = rep2; + rep2 = rep1; + rep1 = rep0; + rep0 = dist; + +Then the decoder decodes exact subtype of "Rep Match" using "IsRepG0", "IsRep0Long", +"IsRepG1", "IsRepG2". + +If the subtype is "Short Rep Match", the decoder updates the state, puts +the one byte from window to current position in window and goes to next +MATCH/LITERAL symbol (the begin of main loop): + + state = UpdateState_ShortRep(state); + OutWindow.PutByte(OutWindow.GetByte(rep0 + 1)); + unpackSize--; + continue; + +In other cases (Rep Match 0/1/2/3), it decodes the zero-based +length of match with "RepLenDecoder" decoder: + + len = RepLenDecoder.Decode(&RangeDec, posState); + +Then it updates the state: + + state = UpdateState_Rep(state); + +Then the decoder must copy match bytes as described in +"The Match symbols copying" section. + + +The match symbols copying +------------------------- + +If we have the match (Simple Match or Rep Match 0/1/2/3), the decoder must +copy the sequence of bytes with calculated match distance and match length. +If uncompressed size is defined, LZMA decoder must check that it doesn't +exceed that specified uncompressed size: + + len += kMatchMinLen; + bool isError = false; + if (unpackSizeDefined && unpackSize < len) + { + len = (unsigned)unpackSize; + isError = true; + } + OutWindow.CopyMatch(rep0 + 1, len); + unpackSize -= len; + if (isError) + return LZMA_RES_ERROR; + +Then the decoder must go to the begin of main loop to decode next MATCH or LITERAL. + + + +NOTES +----- + +This specification doesn't describe the variant of decoder implementation +that supports partial decoding. Such partial decoding case can require some +changes in "end of stream" condition checks code. Also such code +can use additional status codes, returned by decoder. + +This specification uses C++ code with templates to simplify describing. +The optimized version of LZMA decoder doesn't need templates. +Such optimized version can use just two arrays of CProb variables: + 1) The dynamic array of CProb variables allocated for the Literal Decoder. + 2) The one common array that contains all other CProb variables. + + +References: + +1. G. N. N. Martin, Range encoding: an algorithm for removing redundancy + from a digitized message, Video & Data Recording Conference, + Southampton, UK, July 24-27, 1979. diff --git a/deps/LZMA-SDK/DOC/lzma.txt b/deps/LZMA-SDK/DOC/lzma.txt index 1f92142ea..a65988fe0 100644 --- a/deps/LZMA-SDK/DOC/lzma.txt +++ b/deps/LZMA-SDK/DOC/lzma.txt @@ -1,328 +1,328 @@ -LZMA compression ----------------- -Version: 9.35 - -This file describes LZMA encoding and decoding functions written in C language. - -LZMA is an improved version of famous LZ77 compression algorithm. -It was improved in way of maximum increasing of compression ratio, -keeping high decompression speed and low memory requirements for -decompressing. - -Note: you can read also LZMA Specification (lzma-specification.txt from LZMA SDK) - -Also you can look source code for LZMA encoding and decoding: - C/Util/Lzma/LzmaUtil.c - - -LZMA compressed file format ---------------------------- -Offset Size Description - 0 1 Special LZMA properties (lc,lp, pb in encoded form) - 1 4 Dictionary size (little endian) - 5 8 Uncompressed size (little endian). -1 means unknown size - 13 Compressed data - - - -ANSI-C LZMA Decoder -~~~~~~~~~~~~~~~~~~~ - -Please note that interfaces for ANSI-C code were changed in LZMA SDK 4.58. -If you want to use old interfaces you can download previous version of LZMA SDK -from sourceforge.net site. - -To use ANSI-C LZMA Decoder you need the following files: -1) LzmaDec.h + LzmaDec.c + 7zTypes.h + Precomp.h + Compiler.h - -Look example code: - C/Util/Lzma/LzmaUtil.c - - -Memory requirements for LZMA decoding -------------------------------------- - -Stack usage of LZMA decoding function for local variables is not -larger than 200-400 bytes. - -LZMA Decoder uses dictionary buffer and internal state structure. -Internal state structure consumes - state_size = (4 + (1.5 << (lc + lp))) KB -by default (lc=3, lp=0), state_size = 16 KB. - - -How To decompress data ----------------------- - -LZMA Decoder (ANSI-C version) now supports 2 interfaces: -1) Single-call Decompressing -2) Multi-call State Decompressing (zlib-like interface) - -You must use external allocator: -Example: -void *SzAlloc(void *p, size_t size) { p = p; return malloc(size); } -void SzFree(void *p, void *address) { p = p; free(address); } -ISzAlloc alloc = { SzAlloc, SzFree }; - -You can use p = p; operator to disable compiler warnings. - - -Single-call Decompressing -------------------------- -When to use: RAM->RAM decompressing -Compile files: LzmaDec.h + LzmaDec.c + 7zTypes.h -Compile defines: no defines -Memory Requirements: - - Input buffer: compressed size - - Output buffer: uncompressed size - - LZMA Internal Structures: state_size (16 KB for default settings) - -Interface: - int LzmaDecode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, - const Byte *propData, unsigned propSize, ELzmaFinishMode finishMode, - ELzmaStatus *status, ISzAlloc *alloc); - In: - dest - output data - destLen - output data size - src - input data - srcLen - input data size - propData - LZMA properties (5 bytes) - propSize - size of propData buffer (5 bytes) - finishMode - It has meaning only if the decoding reaches output limit (*destLen). - LZMA_FINISH_ANY - Decode just destLen bytes. - LZMA_FINISH_END - Stream must be finished after (*destLen). - You can use LZMA_FINISH_END, when you know that - current output buffer covers last bytes of stream. - alloc - Memory allocator. - - Out: - destLen - processed output size - srcLen - processed input size - - Output: - SZ_OK - status: - LZMA_STATUS_FINISHED_WITH_MARK - LZMA_STATUS_NOT_FINISHED - LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK - SZ_ERROR_DATA - Data error - SZ_ERROR_MEM - Memory allocation error - SZ_ERROR_UNSUPPORTED - Unsupported properties - SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src). - - If LZMA decoder sees end_marker before reaching output limit, it returns OK result, - and output value of destLen will be less than output buffer size limit. - - You can use multiple checks to test data integrity after full decompression: - 1) Check Result and "status" variable. - 2) Check that output(destLen) = uncompressedSize, if you know real uncompressedSize. - 3) Check that output(srcLen) = compressedSize, if you know real compressedSize. - You must use correct finish mode in that case. */ - - -Multi-call State Decompressing (zlib-like interface) ----------------------------------------------------- - -When to use: file->file decompressing -Compile files: LzmaDec.h + LzmaDec.c + 7zTypes.h - -Memory Requirements: - - Buffer for input stream: any size (for example, 16 KB) - - Buffer for output stream: any size (for example, 16 KB) - - LZMA Internal Structures: state_size (16 KB for default settings) - - LZMA dictionary (dictionary size is encoded in LZMA properties header) - -1) read LZMA properties (5 bytes) and uncompressed size (8 bytes, little-endian) to header: - unsigned char header[LZMA_PROPS_SIZE + 8]; - ReadFile(inFile, header, sizeof(header) - -2) Allocate CLzmaDec structures (state + dictionary) using LZMA properties - - CLzmaDec state; - LzmaDec_Constr(&state); - res = LzmaDec_Allocate(&state, header, LZMA_PROPS_SIZE, &g_Alloc); - if (res != SZ_OK) - return res; - -3) Init LzmaDec structure before any new LZMA stream. And call LzmaDec_DecodeToBuf in loop - - LzmaDec_Init(&state); - for (;;) - { - ... - int res = LzmaDec_DecodeToBuf(CLzmaDec *p, Byte *dest, SizeT *destLen, - const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode); - ... - } - - -4) Free all allocated structures - LzmaDec_Free(&state, &g_Alloc); - -Look example code: - C/Util/Lzma/LzmaUtil.c - - -How To compress data --------------------- - -Compile files: - 7zTypes.h - Threads.h - LzmaEnc.h - LzmaEnc.c - LzFind.h - LzFind.c - LzFindMt.h - LzFindMt.c - LzHash.h - -Memory Requirements: - - (dictSize * 11.5 + 6 MB) + state_size - -Lzma Encoder can use two memory allocators: -1) alloc - for small arrays. -2) allocBig - for big arrays. - -For example, you can use Large RAM Pages (2 MB) in allocBig allocator for -better compression speed. Note that Windows has bad implementation for -Large RAM Pages. -It's OK to use same allocator for alloc and allocBig. - - -Single-call Compression with callbacks --------------------------------------- - -Look example code: - C/Util/Lzma/LzmaUtil.c - -When to use: file->file compressing - -1) you must implement callback structures for interfaces: -ISeqInStream -ISeqOutStream -ICompressProgress -ISzAlloc - -static void *SzAlloc(void *p, size_t size) { p = p; return MyAlloc(size); } -static void SzFree(void *p, void *address) { p = p; MyFree(address); } -static ISzAlloc g_Alloc = { SzAlloc, SzFree }; - - CFileSeqInStream inStream; - CFileSeqOutStream outStream; - - inStream.funcTable.Read = MyRead; - inStream.file = inFile; - outStream.funcTable.Write = MyWrite; - outStream.file = outFile; - - -2) Create CLzmaEncHandle object; - - CLzmaEncHandle enc; - - enc = LzmaEnc_Create(&g_Alloc); - if (enc == 0) - return SZ_ERROR_MEM; - - -3) initialize CLzmaEncProps properties; - - LzmaEncProps_Init(&props); - - Then you can change some properties in that structure. - -4) Send LZMA properties to LZMA Encoder - - res = LzmaEnc_SetProps(enc, &props); - -5) Write encoded properties to header - - Byte header[LZMA_PROPS_SIZE + 8]; - size_t headerSize = LZMA_PROPS_SIZE; - UInt64 fileSize; - int i; - - res = LzmaEnc_WriteProperties(enc, header, &headerSize); - fileSize = MyGetFileLength(inFile); - for (i = 0; i < 8; i++) - header[headerSize++] = (Byte)(fileSize >> (8 * i)); - MyWriteFileAndCheck(outFile, header, headerSize) - -6) Call encoding function: - res = LzmaEnc_Encode(enc, &outStream.funcTable, &inStream.funcTable, - NULL, &g_Alloc, &g_Alloc); - -7) Destroy LZMA Encoder Object - LzmaEnc_Destroy(enc, &g_Alloc, &g_Alloc); - - -If callback function return some error code, LzmaEnc_Encode also returns that code -or it can return the code like SZ_ERROR_READ, SZ_ERROR_WRITE or SZ_ERROR_PROGRESS. - - -Single-call RAM->RAM Compression --------------------------------- - -Single-call RAM->RAM Compression is similar to Compression with callbacks, -but you provide pointers to buffers instead of pointers to stream callbacks: - -SRes LzmaEncode(Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, - const CLzmaEncProps *props, Byte *propsEncoded, SizeT *propsSize, int writeEndMark, - ICompressProgress *progress, ISzAlloc *alloc, ISzAlloc *allocBig); - -Return code: - SZ_OK - OK - SZ_ERROR_MEM - Memory allocation error - SZ_ERROR_PARAM - Incorrect paramater - SZ_ERROR_OUTPUT_EOF - output buffer overflow - SZ_ERROR_THREAD - errors in multithreading functions (only for Mt version) - - - -Defines -------- - -_LZMA_SIZE_OPT - Enable some optimizations in LZMA Decoder to get smaller executable code. - -_LZMA_PROB32 - It can increase the speed on some 32-bit CPUs, but memory usage for - some structures will be doubled in that case. - -_LZMA_UINT32_IS_ULONG - Define it if int is 16-bit on your compiler and long is 32-bit. - -_LZMA_NO_SYSTEM_SIZE_T - Define it if you don't want to use size_t type. - - -_7ZIP_PPMD_SUPPPORT - Define it if you don't want to support PPMD method in AMSI-C .7z decoder. - - -C++ LZMA Encoder/Decoder -~~~~~~~~~~~~~~~~~~~~~~~~ -C++ LZMA code use COM-like interfaces. So if you want to use it, -you can study basics of COM/OLE. -C++ LZMA code is just wrapper over ANSI-C code. - - -C++ Notes -~~~~~~~~~~~~~~~~~~~~~~~~ -If you use some C++ code folders in 7-Zip (for example, C++ code for .7z handling), -you must check that you correctly work with "new" operator. -7-Zip can be compiled with MSVC 6.0 that doesn't throw "exception" from "new" operator. -So 7-Zip uses "CPP\Common\NewHandler.cpp" that redefines "new" operator: -operator new(size_t size) -{ - void *p = ::malloc(size); - if (p == 0) - throw CNewException(); - return p; -} -If you use MSCV that throws exception for "new" operator, you can compile without -"NewHandler.cpp". So standard exception will be used. Actually some code of -7-Zip catches any exception in internal code and converts it to HRESULT code. -So you don't need to catch CNewException, if you call COM interfaces of 7-Zip. - ---- - -http://www.7-zip.org -http://www.7-zip.org/sdk.html -http://www.7-zip.org/support.html +LZMA compression +---------------- +Version: 9.35 + +This file describes LZMA encoding and decoding functions written in C language. + +LZMA is an improved version of famous LZ77 compression algorithm. +It was improved in way of maximum increasing of compression ratio, +keeping high decompression speed and low memory requirements for +decompressing. + +Note: you can read also LZMA Specification (lzma-specification.txt from LZMA SDK) + +Also you can look source code for LZMA encoding and decoding: + C/Util/Lzma/LzmaUtil.c + + +LZMA compressed file format +--------------------------- +Offset Size Description + 0 1 Special LZMA properties (lc,lp, pb in encoded form) + 1 4 Dictionary size (little endian) + 5 8 Uncompressed size (little endian). -1 means unknown size + 13 Compressed data + + + +ANSI-C LZMA Decoder +~~~~~~~~~~~~~~~~~~~ + +Please note that interfaces for ANSI-C code were changed in LZMA SDK 4.58. +If you want to use old interfaces you can download previous version of LZMA SDK +from sourceforge.net site. + +To use ANSI-C LZMA Decoder you need the following files: +1) LzmaDec.h + LzmaDec.c + 7zTypes.h + Precomp.h + Compiler.h + +Look example code: + C/Util/Lzma/LzmaUtil.c + + +Memory requirements for LZMA decoding +------------------------------------- + +Stack usage of LZMA decoding function for local variables is not +larger than 200-400 bytes. + +LZMA Decoder uses dictionary buffer and internal state structure. +Internal state structure consumes + state_size = (4 + (1.5 << (lc + lp))) KB +by default (lc=3, lp=0), state_size = 16 KB. + + +How To decompress data +---------------------- + +LZMA Decoder (ANSI-C version) now supports 2 interfaces: +1) Single-call Decompressing +2) Multi-call State Decompressing (zlib-like interface) + +You must use external allocator: +Example: +void *SzAlloc(void *p, size_t size) { p = p; return malloc(size); } +void SzFree(void *p, void *address) { p = p; free(address); } +ISzAlloc alloc = { SzAlloc, SzFree }; + +You can use p = p; operator to disable compiler warnings. + + +Single-call Decompressing +------------------------- +When to use: RAM->RAM decompressing +Compile files: LzmaDec.h + LzmaDec.c + 7zTypes.h +Compile defines: no defines +Memory Requirements: + - Input buffer: compressed size + - Output buffer: uncompressed size + - LZMA Internal Structures: state_size (16 KB for default settings) + +Interface: + int LzmaDecode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + const Byte *propData, unsigned propSize, ELzmaFinishMode finishMode, + ELzmaStatus *status, ISzAlloc *alloc); + In: + dest - output data + destLen - output data size + src - input data + srcLen - input data size + propData - LZMA properties (5 bytes) + propSize - size of propData buffer (5 bytes) + finishMode - It has meaning only if the decoding reaches output limit (*destLen). + LZMA_FINISH_ANY - Decode just destLen bytes. + LZMA_FINISH_END - Stream must be finished after (*destLen). + You can use LZMA_FINISH_END, when you know that + current output buffer covers last bytes of stream. + alloc - Memory allocator. + + Out: + destLen - processed output size + srcLen - processed input size + + Output: + SZ_OK + status: + LZMA_STATUS_FINISHED_WITH_MARK + LZMA_STATUS_NOT_FINISHED + LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK + SZ_ERROR_DATA - Data error + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_UNSUPPORTED - Unsupported properties + SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src). + + If LZMA decoder sees end_marker before reaching output limit, it returns OK result, + and output value of destLen will be less than output buffer size limit. + + You can use multiple checks to test data integrity after full decompression: + 1) Check Result and "status" variable. + 2) Check that output(destLen) = uncompressedSize, if you know real uncompressedSize. + 3) Check that output(srcLen) = compressedSize, if you know real compressedSize. + You must use correct finish mode in that case. */ + + +Multi-call State Decompressing (zlib-like interface) +---------------------------------------------------- + +When to use: file->file decompressing +Compile files: LzmaDec.h + LzmaDec.c + 7zTypes.h + +Memory Requirements: + - Buffer for input stream: any size (for example, 16 KB) + - Buffer for output stream: any size (for example, 16 KB) + - LZMA Internal Structures: state_size (16 KB for default settings) + - LZMA dictionary (dictionary size is encoded in LZMA properties header) + +1) read LZMA properties (5 bytes) and uncompressed size (8 bytes, little-endian) to header: + unsigned char header[LZMA_PROPS_SIZE + 8]; + ReadFile(inFile, header, sizeof(header) + +2) Allocate CLzmaDec structures (state + dictionary) using LZMA properties + + CLzmaDec state; + LzmaDec_Constr(&state); + res = LzmaDec_Allocate(&state, header, LZMA_PROPS_SIZE, &g_Alloc); + if (res != SZ_OK) + return res; + +3) Init LzmaDec structure before any new LZMA stream. And call LzmaDec_DecodeToBuf in loop + + LzmaDec_Init(&state); + for (;;) + { + ... + int res = LzmaDec_DecodeToBuf(CLzmaDec *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode); + ... + } + + +4) Free all allocated structures + LzmaDec_Free(&state, &g_Alloc); + +Look example code: + C/Util/Lzma/LzmaUtil.c + + +How To compress data +-------------------- + +Compile files: + 7zTypes.h + Threads.h + LzmaEnc.h + LzmaEnc.c + LzFind.h + LzFind.c + LzFindMt.h + LzFindMt.c + LzHash.h + +Memory Requirements: + - (dictSize * 11.5 + 6 MB) + state_size + +Lzma Encoder can use two memory allocators: +1) alloc - for small arrays. +2) allocBig - for big arrays. + +For example, you can use Large RAM Pages (2 MB) in allocBig allocator for +better compression speed. Note that Windows has bad implementation for +Large RAM Pages. +It's OK to use same allocator for alloc and allocBig. + + +Single-call Compression with callbacks +-------------------------------------- + +Look example code: + C/Util/Lzma/LzmaUtil.c + +When to use: file->file compressing + +1) you must implement callback structures for interfaces: +ISeqInStream +ISeqOutStream +ICompressProgress +ISzAlloc + +static void *SzAlloc(void *p, size_t size) { p = p; return MyAlloc(size); } +static void SzFree(void *p, void *address) { p = p; MyFree(address); } +static ISzAlloc g_Alloc = { SzAlloc, SzFree }; + + CFileSeqInStream inStream; + CFileSeqOutStream outStream; + + inStream.funcTable.Read = MyRead; + inStream.file = inFile; + outStream.funcTable.Write = MyWrite; + outStream.file = outFile; + + +2) Create CLzmaEncHandle object; + + CLzmaEncHandle enc; + + enc = LzmaEnc_Create(&g_Alloc); + if (enc == 0) + return SZ_ERROR_MEM; + + +3) initialize CLzmaEncProps properties; + + LzmaEncProps_Init(&props); + + Then you can change some properties in that structure. + +4) Send LZMA properties to LZMA Encoder + + res = LzmaEnc_SetProps(enc, &props); + +5) Write encoded properties to header + + Byte header[LZMA_PROPS_SIZE + 8]; + size_t headerSize = LZMA_PROPS_SIZE; + UInt64 fileSize; + int i; + + res = LzmaEnc_WriteProperties(enc, header, &headerSize); + fileSize = MyGetFileLength(inFile); + for (i = 0; i < 8; i++) + header[headerSize++] = (Byte)(fileSize >> (8 * i)); + MyWriteFileAndCheck(outFile, header, headerSize) + +6) Call encoding function: + res = LzmaEnc_Encode(enc, &outStream.funcTable, &inStream.funcTable, + NULL, &g_Alloc, &g_Alloc); + +7) Destroy LZMA Encoder Object + LzmaEnc_Destroy(enc, &g_Alloc, &g_Alloc); + + +If callback function return some error code, LzmaEnc_Encode also returns that code +or it can return the code like SZ_ERROR_READ, SZ_ERROR_WRITE or SZ_ERROR_PROGRESS. + + +Single-call RAM->RAM Compression +-------------------------------- + +Single-call RAM->RAM Compression is similar to Compression with callbacks, +but you provide pointers to buffers instead of pointers to stream callbacks: + +SRes LzmaEncode(Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + const CLzmaEncProps *props, Byte *propsEncoded, SizeT *propsSize, int writeEndMark, + ICompressProgress *progress, ISzAlloc *alloc, ISzAlloc *allocBig); + +Return code: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater + SZ_ERROR_OUTPUT_EOF - output buffer overflow + SZ_ERROR_THREAD - errors in multithreading functions (only for Mt version) + + + +Defines +------- + +_LZMA_SIZE_OPT - Enable some optimizations in LZMA Decoder to get smaller executable code. + +_LZMA_PROB32 - It can increase the speed on some 32-bit CPUs, but memory usage for + some structures will be doubled in that case. + +_LZMA_UINT32_IS_ULONG - Define it if int is 16-bit on your compiler and long is 32-bit. + +_LZMA_NO_SYSTEM_SIZE_T - Define it if you don't want to use size_t type. + + +_7ZIP_PPMD_SUPPPORT - Define it if you don't want to support PPMD method in AMSI-C .7z decoder. + + +C++ LZMA Encoder/Decoder +~~~~~~~~~~~~~~~~~~~~~~~~ +C++ LZMA code use COM-like interfaces. So if you want to use it, +you can study basics of COM/OLE. +C++ LZMA code is just wrapper over ANSI-C code. + + +C++ Notes +~~~~~~~~~~~~~~~~~~~~~~~~ +If you use some C++ code folders in 7-Zip (for example, C++ code for .7z handling), +you must check that you correctly work with "new" operator. +7-Zip can be compiled with MSVC 6.0 that doesn't throw "exception" from "new" operator. +So 7-Zip uses "CPP\Common\NewHandler.cpp" that redefines "new" operator: +operator new(size_t size) +{ + void *p = ::malloc(size); + if (p == 0) + throw CNewException(); + return p; +} +If you use MSCV that throws exception for "new" operator, you can compile without +"NewHandler.cpp". So standard exception will be used. Actually some code of +7-Zip catches any exception in internal code and converts it to HRESULT code. +So you don't need to catch CNewException, if you call COM interfaces of 7-Zip. + +--- + +http://www.7-zip.org +http://www.7-zip.org/sdk.html +http://www.7-zip.org/support.html