Intravision/reactos
1
0
Fork 0
mirror of https://github.com/reactos/reactos.git synced 2025-08-05 21:12:59 +00:00
Code Issues Projects Releases Packages Wiki Activity

[BTRFS][UBTRFS][SHELLBTRFS] Upgrade to 1.7.4 (#4417)

Browse source 
v1.7.4 (2020-08-23):

- Fixed issue when running compressed EXEs
- Changed build system to cmake
- Upgraded zstd to version 1.4.5
- Added support for FSCTL_GET_RETRIEVAL_POINTERS
- Miscellaneous bug fixes
This commit is contained in:
Vincent Franchomme 2022-04-28 21:33:48 +02:00 • committed by Hermès BÉLUSCA - MAÏTO
parent 174dfab667
commit 0604273535
80 changed files with 11313 additions and 6442 deletions
Download patch file Download diff file Expand all files Collapse all files

9
dll/shellext/shellbtrfs/CMakeLists.txt
View file

@ -22,10 +22,17 @@ list(APPEND SOURCE
volpropsheet.cpp)
list(APPEND PCH_SKIP_SOURCE
guid.c)
guid.c
${REACTOS_SOURCE_DIR}/drivers/filesystems/btrfs/crc32c.c)
if((ARCH STREQUAL "i386") OR (ARCH STREQUAL "amd64"))
list(APPEND ASM_SOURCE ${REACTOS_SOURCE_DIR}/drivers/filesystems/btrfs/crc32c.S)
add_asm_files(btrfs_asm ${ASM_SOURCE})
endif()
add_library(shellbtrfs MODULE
${SOURCE}
${btrfs_asm}
${PCH_SKIP_SOURCE}
shellbtrfs.rc
${CMAKE_CURRENT_BINARY_DIR}/shellbtrfs.def)

14
dll/shellext/shellbtrfs/balance.cpp
View file

@ -1061,11 +1061,7 @@ static void unserialize(void* data, ULONG len, WCHAR* s) {
}
}
#ifdef __REACTOS__
extern "C" {
#endif
void CALLBACK StartBalanceW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
extern "C" void CALLBACK StartBalanceW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
try {
WCHAR *s, *vol, *block;
win_handle h, token;
@ -1126,7 +1122,7 @@ void CALLBACK StartBalanceW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int
}
}
void CALLBACK PauseBalanceW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
extern "C" void CALLBACK PauseBalanceW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
try {
win_handle h, token;
TOKEN_PRIVILEGES tp;
@ -1173,7 +1169,7 @@ void CALLBACK PauseBalanceW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int
}
}
void CALLBACK StopBalanceW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
extern "C" void CALLBACK StopBalanceW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
try {
win_handle h, token;
TOKEN_PRIVILEGES tp;
@ -1217,7 +1213,3 @@ void CALLBACK StopBalanceW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int n
error_message(hwnd, e.what());
}
}
#ifdef __REACTOS__
} /* extern "C" */
#endif

10
dll/shellext/shellbtrfs/contextmenu.cpp
View file

@ -1643,11 +1643,7 @@ static void reflink_copy2(const wstring& srcfn, const wstring& destdir, const ws
}
}
#ifdef __REACTOS__
extern "C" {
#endif
void CALLBACK ReflinkCopyW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
extern "C" void CALLBACK ReflinkCopyW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
vector<wstring> args;
command_line_to_args(lpszCmdLine, args);
@ -1720,7 +1716,3 @@ void CALLBACK ReflinkCopyW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int n
}
}
}
#ifdef __REACTOS__
} /* extern "C" */
#endif

4
dll/shellext/shellbtrfs/main.cpp
View file

@ -309,7 +309,7 @@ static void write_reg_key(HKEY root, const wstring& keyname, const WCHAR* val, c
static void register_clsid(const GUID clsid, const WCHAR* description) {
WCHAR* clsidstring;
wstring inproc, progid, clsidkeyname;
wstring inproc, clsidkeyname;
WCHAR dllpath[MAX_PATH];
StringFromCLSID(clsid, &clsidstring);
@ -317,11 +317,9 @@ static void register_clsid(const GUID clsid, const WCHAR* description) {
try {
#ifndef __REACTOS__
inproc = L"CLSID\\"s + clsidstring + L"\\InprocServer32"s;
progid = L"CLSID\\"s + clsidstring + L"\\ProgId"s;
clsidkeyname = L"CLSID\\"s + clsidstring;
#else
inproc = wstring(L"CLSID\\") + clsidstring + wstring(L"\\InprocServer32");
progid = wstring(L"CLSID\\") + clsidstring + wstring(L"\\ProgId");
clsidkeyname = wstring(L"CLSID\\") + clsidstring;
#endif

170
dll/shellext/shellbtrfs/recv.cpp
View file

@ -23,16 +23,15 @@
#include <iostream>
#include "recv.h"
#include "resource.h"
#include "crc32c.h"
#ifndef __REACTOS__
#ifndef _MSC_VER
#include <cpuid.h>
#else
#include <intrin.h>
#endif
#include <smmintrin.h>
#endif
const string EA_NTACL = "security.NTACL";
const string EA_DOSATTRIB = "user.DOSATTRIB";
@ -40,118 +39,6 @@ const string EA_REPARSE = "user.reparse";
const string EA_EA = "user.EA";
const string XATTR_USER = "user.";
#ifndef __REACTOS__
bool have_sse42 = false;
#endif
static const uint32_t crctable[] = {
0x00000000, 0xf26b8303, 0xe13b70f7, 0x1350f3f4, 0xc79a971f, 0x35f1141c, 0x26a1e7e8, 0xd4ca64eb,
0x8ad958cf, 0x78b2dbcc, 0x6be22838, 0x9989ab3b, 0x4d43cfd0, 0xbf284cd3, 0xac78bf27, 0x5e133c24,
0x105ec76f, 0xe235446c, 0xf165b798, 0x030e349b, 0xd7c45070, 0x25afd373, 0x36ff2087, 0xc494a384,
0x9a879fa0, 0x68ec1ca3, 0x7bbcef57, 0x89d76c54, 0x5d1d08bf, 0xaf768bbc, 0xbc267848, 0x4e4dfb4b,
0x20bd8ede, 0xd2d60ddd, 0xc186fe29, 0x33ed7d2a, 0xe72719c1, 0x154c9ac2, 0x061c6936, 0xf477ea35,
0xaa64d611, 0x580f5512, 0x4b5fa6e6, 0xb93425e5, 0x6dfe410e, 0x9f95c20d, 0x8cc531f9, 0x7eaeb2fa,
0x30e349b1, 0xc288cab2, 0xd1d83946, 0x23b3ba45, 0xf779deae, 0x05125dad, 0x1642ae59, 0xe4292d5a,
0xba3a117e, 0x4851927d, 0x5b016189, 0xa96ae28a, 0x7da08661, 0x8fcb0562, 0x9c9bf696, 0x6ef07595,
0x417b1dbc, 0xb3109ebf, 0xa0406d4b, 0x522bee48, 0x86e18aa3, 0x748a09a0, 0x67dafa54, 0x95b17957,
0xcba24573, 0x39c9c670, 0x2a993584, 0xd8f2b687, 0x0c38d26c, 0xfe53516f, 0xed03a29b, 0x1f682198,
0x5125dad3, 0xa34e59d0, 0xb01eaa24, 0x42752927, 0x96bf4dcc, 0x64d4cecf, 0x77843d3b, 0x85efbe38,
0xdbfc821c, 0x2997011f, 0x3ac7f2eb, 0xc8ac71e8, 0x1c661503, 0xee0d9600, 0xfd5d65f4, 0x0f36e6f7,
0x61c69362, 0x93ad1061, 0x80fde395, 0x72966096, 0xa65c047d, 0x5437877e, 0x4767748a, 0xb50cf789,
0xeb1fcbad, 0x197448ae, 0x0a24bb5a, 0xf84f3859, 0x2c855cb2, 0xdeeedfb1, 0xcdbe2c45, 0x3fd5af46,
0x7198540d, 0x83f3d70e, 0x90a324fa, 0x62c8a7f9, 0xb602c312, 0x44694011, 0x5739b3e5, 0xa55230e6,
0xfb410cc2, 0x092a8fc1, 0x1a7a7c35, 0xe811ff36, 0x3cdb9bdd, 0xceb018de, 0xdde0eb2a, 0x2f8b6829,
0x82f63b78, 0x709db87b, 0x63cd4b8f, 0x91a6c88c, 0x456cac67, 0xb7072f64, 0xa457dc90, 0x563c5f93,
0x082f63b7, 0xfa44e0b4, 0xe9141340, 0x1b7f9043, 0xcfb5f4a8, 0x3dde77ab, 0x2e8e845f, 0xdce5075c,
0x92a8fc17, 0x60c37f14, 0x73938ce0, 0x81f80fe3, 0x55326b08, 0xa759e80b, 0xb4091bff, 0x466298fc,
0x1871a4d8, 0xea1a27db, 0xf94ad42f, 0x0b21572c, 0xdfeb33c7, 0x2d80b0c4, 0x3ed04330, 0xccbbc033,
0xa24bb5a6, 0x502036a5, 0x4370c551, 0xb11b4652, 0x65d122b9, 0x97baa1ba, 0x84ea524e, 0x7681d14d,
0x2892ed69, 0xdaf96e6a, 0xc9a99d9e, 0x3bc21e9d, 0xef087a76, 0x1d63f975, 0x0e330a81, 0xfc588982,
0xb21572c9, 0x407ef1ca, 0x532e023e, 0xa145813d, 0x758fe5d6, 0x87e466d5, 0x94b49521, 0x66df1622,
0x38cc2a06, 0xcaa7a905, 0xd9f75af1, 0x2b9cd9f2, 0xff56bd19, 0x0d3d3e1a, 0x1e6dcdee, 0xec064eed,
0xc38d26c4, 0x31e6a5c7, 0x22b65633, 0xd0ddd530, 0x0417b1db, 0xf67c32d8, 0xe52cc12c, 0x1747422f,
0x49547e0b, 0xbb3ffd08, 0xa86f0efc, 0x5a048dff, 0x8ecee914, 0x7ca56a17, 0x6ff599e3, 0x9d9e1ae0,
0xd3d3e1ab, 0x21b862a8, 0x32e8915c, 0xc083125f, 0x144976b4, 0xe622f5b7, 0xf5720643, 0x07198540,
0x590ab964, 0xab613a67, 0xb831c993, 0x4a5a4a90, 0x9e902e7b, 0x6cfbad78, 0x7fab5e8c, 0x8dc0dd8f,
0xe330a81a, 0x115b2b19, 0x020bd8ed, 0xf0605bee, 0x24aa3f05, 0xd6c1bc06, 0xc5914ff2, 0x37faccf1,
0x69e9f0d5, 0x9b8273d6, 0x88d28022, 0x7ab90321, 0xae7367ca, 0x5c18e4c9, 0x4f48173d, 0xbd23943e,
0xf36e6f75, 0x0105ec76, 0x12551f82, 0xe03e9c81, 0x34f4f86a, 0xc69f7b69, 0xd5cf889d, 0x27a40b9e,
0x79b737ba, 0x8bdcb4b9, 0x988c474d, 0x6ae7c44e, 0xbe2da0a5, 0x4c4623a6, 0x5f16d052, 0xad7d5351,
};
// HW code taken from https://github.com/rurban/smhasher/blob/master/crc32_hw.c
#define ALIGN_SIZE 0x08UL
#define ALIGN_MASK (ALIGN_SIZE - 1)
#define CALC_CRC(op, crc, type, buf, len) \
do { \
for (; (len) >= sizeof (type); (len) -= (ULONG)sizeof(type), buf += sizeof (type)) { \
(crc) = (uint32_t)op((crc), *(type *) (buf)); \
} \
} while(0)
#ifndef __REACTOS__
static uint32_t crc32c_hw(const void *input, ULONG len, uint32_t crc) {
const char* buf = (const char*)input;
// Annoyingly, the CRC32 intrinsics don't work properly in modern versions of MSVC -
// it compiles _mm_crc32_u8 as if it was _mm_crc32_u32. And because we're apparently
// not allowed to use inline asm on amd64, there's no easy way to fix this!
for (; (len > 0) && ((size_t)buf & ALIGN_MASK); len--, buf++) {
#ifdef _MSC_VER
crc = crctable[(crc ^ *buf) & 0xff] ^ (crc >> 8);
#else
crc = _mm_crc32_u8(crc, *buf);
#endif
}
#ifdef _AMD64_
#ifdef _MSC_VER
#pragma warning(push)
#pragma warning(disable:4244) // _mm_crc32_u64 wants to return uint64_t(!)
#pragma warning(disable:4242)
#endif
CALC_CRC(_mm_crc32_u64, crc, uint64_t, buf, len);
#ifdef _MSC_VER
#pragma warning(pop)
#endif
#endif
CALC_CRC(_mm_crc32_u32, crc, uint32_t, buf, len);
#ifdef _MSC_VER
for (; len > 0; len--, buf++) {
crc = crctable[(crc ^ *buf) & 0xff] ^ (crc >> 8);
}
#else
CALC_CRC(_mm_crc32_u16, crc, uint16_t, buf, len);
CALC_CRC(_mm_crc32_u8, crc, uint8_t, buf, len);
#endif
return crc;
}
#endif
static uint32_t calc_crc32c(uint32_t seed, uint8_t* msg, ULONG msglen) {
#ifndef __REACTOS__
if (have_sse42)
return crc32c_hw(msg, msglen, seed);
else {
#endif
uint32_t rem;
ULONG i;
rem = seed;
for (i = 0; i < msglen; i++) {
rem = crctable[(rem ^ msg[i]) & 0xff] ^ (rem >> 8);
}
return rem;
#ifndef __REACTOS__
}
#endif
}
bool BtrfsRecv::find_tlv(uint8_t* data, ULONG datalen, uint16_t type, void** value, ULONG* len) {
size_t off = 0;
@ -1473,6 +1360,33 @@ void BtrfsRecv::do_recv(const win_handle& f, uint64_t* pos, uint64_t size, const
}
}
#if defined(_X86_) || defined(_AMD64_)
static void check_cpu() {
bool have_sse42 = false;
#ifndef _MSC_VER
{
uint32_t eax, ebx, ecx, edx;
__cpuid(1, eax, ebx, ecx, edx);
if (__get_cpuid(1, &eax, &ebx, &ecx, &edx))
have_sse42 = ecx & bit_SSE4_2;
}
#else
{
int cpu_info[4];
__cpuid(cpu_info, 1);
have_sse42 = (unsigned int)cpu_info[2] & (1 << 20);
}
#endif
if (have_sse42)
calc_crc32c = calc_crc32c_hw;
}
#endif
DWORD BtrfsRecv::recv_thread() {
LARGE_INTEGER size;
uint64_t pos = 0;
@ -1612,22 +1526,12 @@ static INT_PTR CALLBACK stub_RecvProgressDlgProc(HWND hwndDlg, UINT uMsg, WPARAM
}
void BtrfsRecv::Open(HWND hwnd, const wstring& file, const wstring& path, bool quiet) {
#ifndef __REACTOS__
uint32_t cpuInfo[4];
#endif
streamfile = file;
dirpath = path;
subvolpath = L"";
#ifndef __REACTOS__
#ifndef _MSC_VER
__get_cpuid(1, &cpuInfo[0], &cpuInfo[1], &cpuInfo[2], &cpuInfo[3]);
have_sse42 = cpuInfo[2] & bit_SSE4_2;
#else
__cpuid((int*)cpuInfo, 1);
have_sse42 = cpuInfo[2] & (1 << 20);
#endif
#if defined(_X86_) || defined(_AMD64_)
check_cpu();
#endif
if (quiet)
@ -1638,11 +1542,7 @@ void BtrfsRecv::Open(HWND hwnd, const wstring& file, const wstring& path, bool q
}
}
#ifdef __REACTOS__
extern "C" {
#endif
void CALLBACK RecvSubvolGUIW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
extern "C" void CALLBACK RecvSubvolGUIW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
try {
OPENFILENAMEW ofn;
WCHAR file[MAX_PATH];
@ -1714,7 +1614,7 @@ void CALLBACK RecvSubvolGUIW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int
}
}
void CALLBACK RecvSubvolW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
extern "C" void CALLBACK RecvSubvolW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
try {
vector<wstring> args;
@ -1774,7 +1674,3 @@ void CALLBACK RecvSubvolW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nC
cerr << "Error: " << e.what() << endl;
}
}
#ifdef __REACTOS__
} /* extern "C" */
#endif

14
dll/shellext/shellbtrfs/scrub.cpp
View file

@ -483,11 +483,7 @@ static INT_PTR CALLBACK stub_ScrubDlgProc(HWND hwndDlg, UINT uMsg, WPARAM wParam
return false;
}
#ifdef __REACTOS__
extern "C" {
#endif
void CALLBACK ShowScrubW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
extern "C" void CALLBACK ShowScrubW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
try {
win_handle token;
TOKEN_PRIVILEGES tp;
@ -516,7 +512,7 @@ void CALLBACK ShowScrubW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCm
}
}
void CALLBACK StartScrubW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
extern "C" void CALLBACK StartScrubW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
vector<wstring> args;
command_line_to_args(lpszCmdLine, args);
@ -552,7 +548,7 @@ void CALLBACK StartScrubW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nC
}
}
void CALLBACK StopScrubW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
extern "C" void CALLBACK StopScrubW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
vector<wstring> args;
command_line_to_args(lpszCmdLine, args);
@ -587,7 +583,3 @@ void CALLBACK StopScrubW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCm
}
}
}
#ifdef __REACTOS__
} /* extern "C" */
#endif

20
dll/shellext/shellbtrfs/send.cpp
View file

@ -532,11 +532,7 @@ void BtrfsSend::Open(HWND hwnd, LPWSTR path) {
throw last_error(GetLastError());
}
#ifdef __REACTOS__
extern "C" {
#endif
void CALLBACK SendSubvolGUIW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
extern "C" void CALLBACK SendSubvolGUIW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
try {
win_handle token;
TOKEN_PRIVILEGES tp;
@ -565,10 +561,6 @@ void CALLBACK SendSubvolGUIW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int
}
}
#ifdef __REACTOS__
} /* extern "C" */
#endif
static void send_subvol(const wstring& subvol, const wstring& file, const wstring& parent, const vector<wstring>& clones) {
char* buf;
win_handle dirh, stream;
@ -682,11 +674,7 @@ static void send_subvol(const wstring& subvol, const wstring& file, const wstrin
free(buf);
}
#ifdef __REACTOS__
extern "C" {
#endif
void CALLBACK SendSubvolW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
extern "C" void CALLBACK SendSubvolW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nCmdShow) {
vector<wstring> args;
wstring subvol = L"", parent = L"", file = L"";
vector<wstring> clones;
@ -742,7 +730,3 @@ void CALLBACK SendSubvolW(HWND hwnd, HINSTANCE hinst, LPWSTR lpszCmdLine, int nC
}
}
}
#ifdef __REACTOS__
} /* extern "C" */
#endif

8
dll/shellext/shellbtrfs/shellbtrfs.rc
View file

@ -61,8 +61,8 @@ IDI_ICON1 ICON "subvol.ico"
//
VS_VERSION_INFO VERSIONINFO
FILEVERSION 1,7,3,0
PRODUCTVERSION 1,7,3,0
FILEVERSION 1,7,4,0
PRODUCTVERSION 1,7,4,0
FILEFLAGSMASK 0x17L
#ifdef _DEBUG
FILEFLAGS 0x1L
@ -78,12 +78,12 @@ BEGIN
BLOCK "080904b0"
BEGIN
VALUE "FileDescription", "WinBtrfs shell extension"
VALUE "FileVersion", "1.7.3"
VALUE "FileVersion", "1.7.4"
VALUE "InternalName", "btrfs"
VALUE "LegalCopyright", "Copyright (c) Mark Harmstone 2016-20"
VALUE "OriginalFilename", "shellbtrfs.dll"
VALUE "ProductName", "WinBtrfs"
VALUE "ProductVersion", "1.7.3"
VALUE "ProductVersion", "1.7.4"
END
END
BLOCK "VarFileInfo"

21
dll/shellext/shellbtrfs/volpropsheet.cpp
View file

@ -1590,9 +1590,7 @@ INT_PTR BtrfsChangeDriveLetter::DlgProc(HWND hwndDlg, UINT uMsg, WPARAM wParam,
return false;
}
#ifdef __REACTOS__
INT_PTR CALLBACK VolPropSheetDlgproc(HWND hwndDlg, UINT uMsg, WPARAM wParam, LPARAM lParam)
{
static INT_PTR __stdcall dlg_proc(HWND hwndDlg, UINT uMsg, WPARAM wParam, LPARAM lParam) {
BtrfsChangeDriveLetter* bcdl;
if (uMsg == WM_INITDIALOG) {
@ -1603,24 +1601,9 @@ INT_PTR CALLBACK VolPropSheetDlgproc(HWND hwndDlg, UINT uMsg, WPARAM wParam, LPA
return bcdl->DlgProc(hwndDlg, uMsg, wParam, lParam);
}
#endif
void BtrfsChangeDriveLetter::show() {
#ifndef __REACTOS__
DialogBoxParamW(module, MAKEINTRESOURCEW(IDD_DRIVE_LETTER), hwnd, [](HWND hwndDlg, UINT uMsg, WPARAM wParam, LPARAM lParam) {
BtrfsChangeDriveLetter* bcdl;
if (uMsg == WM_INITDIALOG) {
SetWindowLongPtr(hwndDlg, GWLP_USERDATA, (LONG_PTR)lParam);
bcdl = (BtrfsChangeDriveLetter*)lParam;
} else
bcdl = (BtrfsChangeDriveLetter*)GetWindowLongPtr(hwndDlg, GWLP_USERDATA);
return bcdl->DlgProc(hwndDlg, uMsg, wParam, lParam);
}, (LPARAM)this);
#else
DialogBoxParamW(module, MAKEINTRESOURCEW(IDD_DRIVE_LETTER), hwnd, VolPropSheetDlgproc, (LPARAM)this);
#endif
DialogBoxParamW(module, MAKEINTRESOURCEW(IDD_DRIVE_LETTER), hwnd, dlg_proc, (LPARAM)this);
}
#ifdef __cplusplus

8
dll/win32/ubtrfs/ubtrfs.rc
View file

@ -51,8 +51,8 @@ END
//
VS_VERSION_INFO VERSIONINFO
FILEVERSION 1,7,3,0
PRODUCTVERSION 1,7,3,0
FILEVERSION 1,7,4,0
PRODUCTVERSION 1,7,4,0
FILEFLAGSMASK 0x17L
#ifdef _DEBUG
FILEFLAGS 0x1L
@ -68,12 +68,12 @@ BEGIN
BLOCK "080904b0"
BEGIN
VALUE "FileDescription", "Btrfs utility DLL"
VALUE "FileVersion", "1.7.3"
VALUE "FileVersion", "1.7.4"
VALUE "InternalName", "ubtrfs"
VALUE "LegalCopyright", "Copyright (c) Mark Harmstone 2016-20"
VALUE "OriginalFilename", "ubtrfs.dll"
VALUE "ProductName", "WinBtrfs"
VALUE "ProductVersion", "1.7.3"
VALUE "ProductVersion", "1.7.4"
END
END
BLOCK "VarFileInfo"

47
drivers/filesystems/btrfs/CMakeLists.txt
View file

@ -3,6 +3,28 @@ include_directories(${REACTOS_SOURCE_DIR}/sdk/include/reactos/drivers
${REACTOS_SOURCE_DIR}/sdk/include/reactos/libs/zlib
inc)
set(ZSTD_SRC_FILES
zstd/entropy_common.c
zstd/error_private.c
zstd/fse_compress.c
zstd/fse_decompress.c
zstd/hist.c
zstd/huf_compress.c
zstd/huf_decompress.c
zstd/zstd_common.c
zstd/zstd_compress.c
zstd/zstd_compress_literals.c
zstd/zstd_compress_sequences.c
zstd/zstd_compress_superblock.c
zstd/zstd_ddict.c
zstd/zstd_decompress.c
zstd/zstd_decompress_block.c
zstd/zstd_double_fast.c
zstd/zstd_fast.c
zstd/zstd_lazy.c
zstd/zstd_ldm.c
zstd/zstd_opt.c)
list(APPEND SOURCE
balance.c
blake2b-ref.c
@ -38,31 +60,14 @@ list(APPEND SOURCE
worker-thread.c
write.c
xxhash.c
zstd/entropy_common.c
zstd/fse_compress.c
zstd/hist.c
zstd/huf_decompress.c
zstd/zstd_common.c
zstd/zstd_decompress.c
zstd/zstd_fast.c
zstd/zstd_ldm.c
zstd/error_private.c
zstd/fse_decompress.c
zstd/huf_compress.c
zstd/zstd_compress.c
zstd/zstd_double_fast.c
zstd/zstd_lazy.c
zstd/zstd_opt.c
${ZSTD_SRC_FILES}
btrfs_drv.h)
if(ARCH STREQUAL "i386")
list(APPEND ASM_SOURCE crc32c-x86.S)
elseif(ARCH STREQUAL "amd64")
list(APPEND ASM_SOURCE crc32c-amd64.S)
if((ARCH STREQUAL "i386") OR (ARCH STREQUAL "amd64"))
list(APPEND ASM_SOURCE crc32c.S xor.S)
add_asm_files(btrfs_asm ${ASM_SOURCE})
endif()
add_asm_files(btrfs_asm ${ASM_SOURCE})
add_library(btrfs MODULE ${SOURCE} ${btrfs_asm} btrfs.rc)
if(MSVC)

4
drivers/filesystems/btrfs/boot.c
View file

@ -506,6 +506,10 @@ void __stdcall check_system_root(PDRIVER_OBJECT DriverObject, PVOID Context, ULO
TRACE("(%p, %p, %lu)\n", DriverObject, Context, Count);
UNUSED(DriverObject);
UNUSED(Context);
UNUSED(Count);
// wait for any PNP notifications in progress to finish
ExAcquireResourceExclusiveLite(&boot_lock, TRUE);
ExReleaseResourceLite(&boot_lock);

143
drivers/filesystems/btrfs/btrfs.c
View file

@ -63,9 +63,6 @@ DEFINE_GUID(BtrfsBusInterface, 0x4d414874, 0x6865, 0x6761, 0x6d, 0x65, 0x83, 0x6
PDRIVER_OBJECT drvobj;
PDEVICE_OBJECT master_devobj, busobj;
#ifndef __REACTOS__
bool have_sse2 = false;
#endif
uint64_t num_reads = 0;
LIST_ENTRY uid_map_list, gid_map_list;
LIST_ENTRY VcbList;
@ -123,6 +120,9 @@ static void init_serial(bool first_time);
#endif
static NTSTATUS close_file(_In_ PFILE_OBJECT FileObject, _In_ PIRP Irp);
static void __stdcall do_xor_basic(uint8_t* buf1, uint8_t* buf2, uint32_t len);
xor_func do_xor = do_xor_basic;
typedef struct {
KEVENT Event;
@ -282,6 +282,49 @@ bool is_top_level(_In_ PIRP Irp) {
return false;
}
static void __stdcall do_xor_basic(uint8_t* buf1, uint8_t* buf2, uint32_t len) {
uint32_t j;
#if defined(_ARM_) || defined(_ARM64_)
uint64x2_t x1, x2;
if (((uintptr_t)buf1 & 0xf) == 0 && ((uintptr_t)buf2 & 0xf) == 0) {
while (len >= 16) {
x1 = vld1q_u64((const uint64_t*)buf1);
x2 = vld1q_u64((const uint64_t*)buf2);
x1 = veorq_u64(x1, x2);
vst1q_u64((uint64_t*)buf1, x1);
buf1 += 16;
buf2 += 16;
len -= 16;
}
}
#endif
#if defined(_AMD64_) || defined(_ARM64_)
while (len > 8) {
*(uint64_t*)buf1 ^= *(uint64_t*)buf2;
buf1 += 8;
buf2 += 8;
len -= 8;
}
#endif
while (len > 4) {
*(uint32_t*)buf1 ^= *(uint32_t*)buf2;
buf1 += 4;
buf2 += 4;
len -= 4;
}
for (j = 0; j < len; j++) {
*buf1 ^= *buf2;
buf1++;
buf2++;
}
}
_Function_class_(DRIVER_UNLOAD)
static void __stdcall DriverUnload(_In_ PDRIVER_OBJECT DriverObject) {
UNICODE_STRING dosdevice_nameW;
@ -520,7 +563,7 @@ static NTSTATUS __stdcall drv_flush_buffers(_In_ PDEVICE_OBJECT DeviceObject, _I
top_level = is_top_level(Irp);
if (Vcb && Vcb->type == VCB_TYPE_VOLUME) {
Status = vol_flush_buffers(DeviceObject, Irp);
Status = STATUS_SUCCESS;
goto end;
} else if (!Vcb || Vcb->type != VCB_TYPE_FS) {
Status = STATUS_SUCCESS;
@ -726,14 +769,8 @@ NTSTATUS utf8_to_utf16(WCHAR* dest, ULONG dest_max, ULONG* dest_len, char* src,
uint8_t* in = (uint8_t*)src;
uint16_t* out = (uint16_t*)dest;
ULONG needed = 0, left = dest_max / sizeof(uint16_t);
#ifdef __REACTOS__
ULONG i;
for (i = 0; i < src_len; ++i) {
#else
for (ULONG i = 0; i < src_len; i++) {
#endif
uint32_t cp;
if (!(in[i] & 0x80))
@ -816,14 +853,8 @@ NTSTATUS utf16_to_utf8(char* dest, ULONG dest_max, ULONG* dest_len, WCHAR* src,
uint8_t* out = (uint8_t*)dest;
ULONG in_len = src_len / sizeof(uint16_t);
ULONG needed = 0, left = dest_max;
#ifdef __REACTOS__
ULONG i = 0;
for (i = 0; i < in_len; i++) {
#else
for (ULONG i = 0; i < in_len; i++) {
#endif
uint32_t cp = *in;
in++;
@ -934,7 +965,7 @@ static NTSTATUS __stdcall drv_query_volume_information(_In_ PDEVICE_OBJECT Devic
top_level = is_top_level(Irp);
if (Vcb && Vcb->type == VCB_TYPE_VOLUME) {
Status = vol_query_volume_information(DeviceObject, Irp);
Status = STATUS_INVALID_DEVICE_REQUEST;
goto end;
} else if (!Vcb || Vcb->type != VCB_TYPE_FS) {
Status = STATUS_INVALID_PARAMETER;
@ -1091,9 +1122,9 @@ static NTSTATUS __stdcall drv_query_volume_information(_In_ PDEVICE_OBJECT Devic
orig_label_len = label_len;
if (IrpSp->Parameters.QueryVolume.Length < sizeof(FILE_FS_VOLUME_INFORMATION) - sizeof(WCHAR) + label_len) {
if (IrpSp->Parameters.QueryVolume.Length > sizeof(FILE_FS_VOLUME_INFORMATION) - sizeof(WCHAR))
label_len = IrpSp->Parameters.QueryVolume.Length - sizeof(FILE_FS_VOLUME_INFORMATION) + sizeof(WCHAR);
if (IrpSp->Parameters.QueryVolume.Length < offsetof(FILE_FS_VOLUME_INFORMATION, VolumeLabel) + label_len) {
if (IrpSp->Parameters.QueryVolume.Length > offsetof(FILE_FS_VOLUME_INFORMATION, VolumeLabel))
label_len = IrpSp->Parameters.QueryVolume.Length - offsetof(FILE_FS_VOLUME_INFORMATION, VolumeLabel);
else
label_len = 0;
@ -1102,12 +1133,12 @@ static NTSTATUS __stdcall drv_query_volume_information(_In_ PDEVICE_OBJECT Devic
TRACE("label_len = %lu\n", label_len);
ffvi.VolumeCreationTime.QuadPart = 0; // FIXME
RtlZeroMemory(&ffvi, offsetof(FILE_FS_VOLUME_INFORMATION, VolumeLabel));
ffvi.VolumeSerialNumber = Vcb->superblock.uuid.uuid[12] << 24 | Vcb->superblock.uuid.uuid[13] << 16 | Vcb->superblock.uuid.uuid[14] << 8 | Vcb->superblock.uuid.uuid[15];
ffvi.VolumeLabelLength = orig_label_len;
ffvi.SupportsObjects = false;
RtlCopyMemory(data, &ffvi, min(sizeof(FILE_FS_VOLUME_INFORMATION) - sizeof(WCHAR), IrpSp->Parameters.QueryVolume.Length));
RtlCopyMemory(data, &ffvi, min(offsetof(FILE_FS_VOLUME_INFORMATION, VolumeLabel), IrpSp->Parameters.QueryVolume.Length));
if (label_len > 0) {
ULONG bytecount;
@ -1124,7 +1155,7 @@ static NTSTATUS __stdcall drv_query_volume_information(_In_ PDEVICE_OBJECT Devic
ExReleaseResourceLite(&Vcb->tree_lock);
BytesCopied = sizeof(FILE_FS_VOLUME_INFORMATION) - sizeof(WCHAR) + label_len;
BytesCopied = offsetof(FILE_FS_VOLUME_INFORMATION, VolumeLabel) + label_len;
Status = overflow ? STATUS_BUFFER_OVERFLOW : STATUS_SUCCESS;
break;
}
@ -1384,7 +1415,7 @@ static NTSTATUS __stdcall drv_set_volume_information(_In_ PDEVICE_OBJECT DeviceO
top_level = is_top_level(Irp);
if (Vcb && Vcb->type == VCB_TYPE_VOLUME) {
Status = vol_set_volume_information(DeviceObject, Irp);
Status = STATUS_INVALID_DEVICE_REQUEST;
goto end;
} else if (!Vcb || Vcb->type != VCB_TYPE_FS) {
Status = STATUS_INVALID_PARAMETER;
@ -1779,12 +1810,8 @@ void reap_fcbs(device_extension* Vcb) {
}
void free_fileref(_Inout_ file_ref* fr) {
LONG rc;
rc = InterlockedDecrement(&fr->refcount);
#ifdef __REACTOS__
(void)rc;
#endif
#if defined(_DEBUG) || defined(DEBUG_FCB_REFCOUNTS)
LONG rc = InterlockedDecrement(&fr->refcount);
#ifdef DEBUG_FCB_REFCOUNTS
ERR("fileref %p: refcount now %i\n", fr, rc);
@ -1796,6 +1823,9 @@ void free_fileref(_Inout_ file_ref* fr) {
int3;
}
#endif
#else
InterlockedDecrement(&fr->refcount);
#endif
}
void reap_fileref(device_extension* Vcb, file_ref* fr) {
@ -2383,7 +2413,8 @@ static NTSTATUS __stdcall drv_cleanup(_In_ PDEVICE_OBJECT DeviceObject, _In_ PIR
top_level = is_top_level(Irp);
if (Vcb && Vcb->type == VCB_TYPE_VOLUME) {
Status = vol_cleanup(DeviceObject, Irp);
Irp->IoStatus.Information = 0;
Status = STATUS_SUCCESS;
goto exit;
} else if (DeviceObject == master_devobj) {
TRACE("closing file system\n");
@ -5215,7 +5246,7 @@ static NTSTATUS __stdcall drv_file_system_control(_In_ PDEVICE_OBJECT DeviceObje
top_level = is_top_level(Irp);
if (Vcb && Vcb->type == VCB_TYPE_VOLUME) {
Status = vol_file_system_control(DeviceObject, Irp);
Status = STATUS_INVALID_DEVICE_REQUEST;
goto end;
} else if (!Vcb || (Vcb->type != VCB_TYPE_FS && Vcb->type != VCB_TYPE_CONTROL)) {
Status = STATUS_INVALID_PARAMETER;
@ -5295,7 +5326,7 @@ static NTSTATUS __stdcall drv_lock_control(_In_ PDEVICE_OBJECT DeviceObject, _In
top_level = is_top_level(Irp);
if (Vcb && Vcb->type == VCB_TYPE_VOLUME) {
Status = vol_lock_control(DeviceObject, Irp);
Status = STATUS_INVALID_DEVICE_REQUEST;
Irp->IoStatus.Status = Status;
IoCompleteRequest(Irp, IO_NO_INCREMENT);
@ -5460,7 +5491,7 @@ static NTSTATUS __stdcall drv_shutdown(_In_ PDEVICE_OBJECT DeviceObject, _In_ PI
top_level = is_top_level(Irp);
if (Vcb && Vcb->type == VCB_TYPE_VOLUME) {
Status = vol_shutdown(DeviceObject, Irp);
Status = STATUS_INVALID_DEVICE_REQUEST;
goto end;
}
@ -5888,16 +5919,29 @@ static void init_serial(bool first_time) {
#if !defined(__REACTOS__) && (defined(_X86_) || defined(_AMD64_))
static void check_cpu() {
bool have_sse42;
bool have_sse2 = false, have_sse42 = false, have_avx2 = false;
#ifndef _MSC_VER
{
uint32_t eax, ebx, ecx, edx;
uint32_t eax, ebx, ecx, edx, xcr0;
__cpuid(1, eax, ebx, ecx, edx);
if (__get_cpuid(1, &eax, &ebx, &ecx, &edx)) {
have_sse42 = ecx & bit_SSE4_2;
have_sse2 = edx & bit_SSE2;
}
if (__get_cpuid_count(7, 0, &eax, &ebx, &ecx, &edx))
have_avx2 = ebx & bit_AVX2;
if (have_avx2) { // check if supported by OS
__asm__("xgetbv" : "=a" (xcr0) : "c" (0) : "edx" );
if ((xcr0 & 6) != 6)
have_avx2 = false;
}
}
#else
{
unsigned int cpu_info[4];
@ -5905,6 +5949,16 @@ static void check_cpu() {
__cpuid(cpu_info, 1);
have_sse42 = cpu_info[2] & (1 << 20);
have_sse2 = cpu_info[3] & (1 << 26);
__cpuidex(cpu_info, 7, 0);
have_avx2 = cpu_info[1] & (1 << 5);
if (have_avx2) {
uint32_t xcr0 = (uint32_t)_xgetbv(0);
if ((xcr0 & 6) != 6)
have_avx2 = false;
}
}
#endif
@ -5914,10 +5968,19 @@ static void check_cpu() {
} else
TRACE("SSE4.2 not supported\n");
if (have_sse2)
if (have_sse2) {
TRACE("SSE2 is supported\n");
else
if (!have_avx2)
do_xor = do_xor_sse2;
} else
TRACE("SSE2 is not supported\n");
if (have_avx2) {
TRACE("AVX2 is supported\n");
do_xor = do_xor_avx2;
} else
TRACE("AVX2 is not supported\n");
}
#endif
@ -6047,6 +6110,8 @@ NTSTATUS __stdcall AddDevice(PDRIVER_OBJECT DriverObject, PDEVICE_OBJECT Physica
TRACE("(%p, %p)\n", DriverObject, PhysicalDeviceObject);
UNUSED(DriverObject);
ExAcquireResourceSharedLite(&pdo_list_lock, true);
le = pdo_list.Flink;

8
drivers/filesystems/btrfs/btrfs.rc
View file

@ -51,8 +51,8 @@ END
//
VS_VERSION_INFO VERSIONINFO
FILEVERSION 1,7,3,0
PRODUCTVERSION 1,7,3,0
FILEVERSION 1,7,4,0
PRODUCTVERSION 1,7,4,0
FILEFLAGSMASK 0x17L
#ifdef _DEBUG
FILEFLAGS 0x1L
@ -68,12 +68,12 @@ BEGIN
BLOCK "080904b0"
BEGIN
VALUE "FileDescription", "WinBtrfs"
VALUE "FileVersion", "1.7.3"
VALUE "FileVersion", "1.7.4"
VALUE "InternalName", "btrfs"
VALUE "LegalCopyright", "Copyright (c) Mark Harmstone 2016-20"
VALUE "OriginalFilename", "btrfs.sys"
VALUE "ProductName", "WinBtrfs"
VALUE "ProductVersion", "1.7.3"
VALUE "ProductVersion", "1.7.4"
END
END
BLOCK "VarFileInfo"

87
drivers/filesystems/btrfs/btrfs_drv.h
View file

@ -66,10 +66,6 @@
#include "btrfs.h"
#include "btrfsioctl.h"
#if !defined(__REACTOS__) && (defined(_X86_) || defined(_AMD64_))
#include <emmintrin.h>
#endif
#ifdef __REACTOS__
C_ASSERT(sizeof(bool) == 1);
#endif
@ -134,14 +130,16 @@ C_ASSERT(sizeof(bool) == 1);
#define try __try
#define except __except
#define finally __finally
#define leave __leave
#else
#define try if (1)
#define except(x) if (0 && (x))
#define finally if (1)
#define leave
#endif
#ifndef __REACTOS__
#ifdef __GNUC__
#ifndef InterlockedIncrement64
#define InterlockedIncrement64(a) __sync_add_and_fetch(a, 1)
#endif
#endif // __REACTOS__
@ -1102,6 +1100,12 @@ __inline static uint32_t get_extent_data_refcount(uint8_t type, void* data) {
}
}
// in xor-gas.S
#if defined(_X86_) || defined(_AMD64_)
void __stdcall do_xor_sse2(uint8_t* buf1, uint8_t* buf2, uint32_t len);
void __stdcall do_xor_avx2(uint8_t* buf1, uint8_t* buf2, uint32_t len);
#endif
// in btrfs.c
_Ret_maybenull_
device* find_device_from_uuid(_In_ device_extension* Vcb, _In_ BTRFS_UUID* uuid);
@ -1131,6 +1135,10 @@ bool is_file_name_valid(_In_ PUNICODE_STRING us, _In_ bool posix, _In_ bool stre
void send_notification_fileref(_In_ file_ref* fileref, _In_ ULONG filter_match, _In_ ULONG action, _In_opt_ PUNICODE_STRING stream);
void queue_notification_fcb(_In_ file_ref* fileref, _In_ ULONG filter_match, _In_ ULONG action, _In_opt_ PUNICODE_STRING stream);
typedef void (__stdcall *xor_func)(uint8_t* buf1, uint8_t* buf2, uint32_t len);
extern xor_func do_xor;
#ifdef DEBUG_CHUNK_LOCKS
#define acquire_chunk_lock(c, Vcb) { ExAcquireResourceExclusiveLite(&c->lock, true); InterlockedIncrement(&Vcb->chunk_locks_held); }
#define release_chunk_lock(c, Vcb) { InterlockedDecrement(&Vcb->chunk_locks_held); ExReleaseResourceLite(&c->lock); }
@ -1176,8 +1184,6 @@ bool check_superblock_checksum(superblock* sb);
#define funcname __func__
#endif
extern bool have_sse2;
extern uint32_t mount_compress;
extern uint32_t mount_compress_force;
extern uint32_t mount_compress_type;
@ -1229,8 +1235,8 @@ void _debug_message(_In_ const char* func, _In_ char* s, ...) __attribute__((for
#else
#define TRACE(s, ...)
#define WARN(s, ...)
#define TRACE(s, ...) do { } while(0)
#define WARN(s, ...) do { } while(0)
#define FIXME(s, ...) DbgPrint("Btrfs FIXME : %s : " s, funcname, ##__VA_ARGS__)
#define ERR(s, ...) DbgPrint("Btrfs ERR : %s : " s, funcname, ##__VA_ARGS__)
@ -1424,10 +1430,10 @@ void insert_dir_child_into_hash_lists(fcb* fcb, dir_child* dc);
void remove_dir_child_from_hash_lists(fcb* fcb, dir_child* dc);
// in reparse.c
NTSTATUS get_reparse_point(PDEVICE_OBJECT DeviceObject, PFILE_OBJECT FileObject, void* buffer, DWORD buflen, ULONG_PTR* retlen);
NTSTATUS get_reparse_point(PFILE_OBJECT FileObject, void* buffer, DWORD buflen, ULONG_PTR* retlen);
NTSTATUS set_reparse_point2(fcb* fcb, REPARSE_DATA_BUFFER* rdb, ULONG buflen, ccb* ccb, file_ref* fileref, PIRP Irp, LIST_ENTRY* rollback);
NTSTATUS set_reparse_point(PDEVICE_OBJECT DeviceObject, PIRP Irp);
NTSTATUS delete_reparse_point(PDEVICE_OBJECT DeviceObject, PIRP Irp);
NTSTATUS set_reparse_point(PIRP Irp);
NTSTATUS delete_reparse_point(PIRP Irp);
// in create.c
@ -1605,21 +1611,7 @@ NTSTATUS vol_create(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_close(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_read(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_write(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_query_information(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_set_information(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_query_ea(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_set_ea(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_flush_buffers(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_query_volume_information(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_set_volume_information(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_cleanup(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_directory_control(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_file_system_control(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_lock_control(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_device_control(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_shutdown(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_query_security(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
NTSTATUS vol_set_security(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp);
void add_volume_device(superblock* sb, PUNICODE_STRING devpath, uint64_t length, ULONG disk_num, ULONG part_num);
NTSTATUS mountmgr_add_drive_letter(PDEVICE_OBJECT mountmgr, PUNICODE_STRING devpath);
@ -1734,49 +1726,6 @@ static __inline bool write_fcb_compressed(fcb* fcb) {
return false;
}
static __inline void do_xor(uint8_t* buf1, uint8_t* buf2, uint32_t len) {
uint32_t j;
#ifndef __REACTOS__
#if defined(_X86_) || defined(_AMD64_)
__m128i x1, x2;
if (have_sse2 && ((uintptr_t)buf1 & 0xf) == 0 && ((uintptr_t)buf2 & 0xf) == 0) {
while (len >= 16) {
x1 = _mm_load_si128((__m128i*)buf1);
x2 = _mm_load_si128((__m128i*)buf2);
x1 = _mm_xor_si128(x1, x2);
_mm_store_si128((__m128i*)buf1, x1);
buf1 += 16;
buf2 += 16;
len -= 16;
}
}
#elif defined(_ARM_) || defined(_ARM64_)
uint64x2_t x1, x2;
if (((uintptr_t)buf1 & 0xf) == 0 && ((uintptr_t)buf2 & 0xf) == 0) {
while (len >= 16) {
x1 = vld1q_u64((const uint64_t*)buf1);
x2 = vld1q_u64((const uint64_t*)buf2);
x1 = veorq_u64(x1, x2);
vst1q_u64((uint64_t*)buf1, x1);
buf1 += 16;
buf2 += 16;
len -= 16;
}
}
#endif
#endif // __REACTOS__
for (j = 0; j < len; j++) {
*buf1 ^= *buf2;
buf1++;
buf2++;
}
}
#ifdef DEBUG_FCB_REFCOUNTS
#ifdef DEBUG_LONG_MESSAGES
#define increase_fileref_refcount(fileref) {\

8
drivers/filesystems/btrfs/btrfsioctl.h
View file

@ -37,6 +37,7 @@
#define FSCTL_BTRFS_READ_SEND_BUFFER CTL_CODE(FILE_DEVICE_UNKNOWN, 0x847, METHOD_OUT_DIRECT, FILE_ANY_ACCESS)
#define FSCTL_BTRFS_RESIZE CTL_CODE(FILE_DEVICE_UNKNOWN, 0x848, METHOD_IN_DIRECT, FILE_ANY_ACCESS)
#define IOCTL_BTRFS_UNLOAD CTL_CODE(FILE_DEVICE_UNKNOWN, 0x849, METHOD_NEITHER, FILE_ANY_ACCESS)
#define FSCTL_BTRFS_GET_CSUM_INFO CTL_CODE(FILE_DEVICE_UNKNOWN, 0x84a, METHOD_BUFFERED, FILE_READ_ACCESS)
typedef struct {
uint64_t subvol;
@ -279,3 +280,10 @@ typedef struct {
uint64_t device;
uint64_t size;
} btrfs_resize;
typedef struct {
uint8_t csum_type;
uint8_t csum_length;
uint64_t num_sectors;
uint8_t data[1];
} btrfs_csum_info;

5
drivers/filesystems/btrfs/compress.c
View file

@ -703,6 +703,7 @@ NTSTATUS zstd_decompress(uint8_t* inbuf, uint32_t inlen, uint8_t* outbuf, uint32
output.size = outlen;
output.pos = 0;
do {
read = ZSTD_decompressStream(stream, &output, &input);
if (ZSTD_isError(read)) {
@ -711,6 +712,10 @@ NTSTATUS zstd_decompress(uint8_t* inbuf, uint32_t inlen, uint8_t* outbuf, uint32
goto end;
}
if (output.pos == output.size)
break;
} while (read != 0);
Status = STATUS_SUCCESS;
end:

113
drivers/filesystems/btrfs/crc32c-amd64.S
View file

@ -1,113 +0,0 @@
/* Copyright (c) Mark Harmstone 2020
*
* This file is part of WinBtrfs.
*
* WinBtrfs is free software: you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public Licence as published by
* the Free Software Foundation, either version 3 of the Licence, or
* (at your option) any later version.
*
* WinBtrfs is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Lesser General Public Licence for more details.
*
* You should have received a copy of the GNU Lesser General Public Licence
* along with WinBtrfs. If not, see <http://www.gnu.org/licenses/>. */
#include <asm.inc>
EXTERN crctable:QWORD
.code64
/* uint32_t __stdcall calc_crc32c_sw(uint32_t seed, uint8_t* msg, uint32_t msglen); */
PUBLIC calc_crc32c_sw
calc_crc32c_sw:
/* rax = crc / seed
* rdx = buf
* r8 = len
* rcx = tmp
* r10 = tmp2 */
mov rax, rcx
crcloop:
test r8, r8
jz crcend
mov rcx, rax
shr rcx, 8
mov r10b, byte ptr [rdx]
xor al, r10b
and rax, 255
shl rax, 2
lea r10, [rip+crctable]
mov eax, dword ptr [r10 + rax]
xor rax, rcx
inc rdx
dec r8
jmp crcloop
crcend:
ret
/****************************************************/
/* uint32_t __stdcall calc_crc32c_hw(uint32_t seed, uint8_t* msg, uint32_t msglen); */
PUBLIC calc_crc32c_hw
calc_crc32c_hw:
/* rax = crc / seed
* rdx = buf
* r8 = len */
mov rax, rcx
crchw_loop:
cmp r8, 8
jl crchw_stragglers
crc32 rax, qword ptr [rdx]
add rdx, 8
sub r8, 8
jmp crchw_loop
crchw_stragglers:
cmp r8, 4
jl crchw_stragglers2
crc32 eax, dword ptr [rdx]
add rdx, 4
sub r8, 4
crchw_stragglers2:
cmp r8, 2
jl crchw_stragglers3
crc32 eax, word ptr [rdx]
add rdx, 2
sub r8, 2
crchw_stragglers3:
test r8, r8
jz crchw_end
crc32 eax, byte ptr [rdx]
inc rdx
dec r8
jmp crchw_stragglers3
crchw_end:
ret
END

100
drivers/filesystems/btrfs/crc32c-x86.S → drivers/filesystems/btrfs/crc32c.S
View file

@ -17,8 +17,105 @@
#include <asm.inc>
EXTERN _crctable:DWORD
#ifdef __x86_64__
EXTERN crctable:QWORD
.code64
/* uint32_t __stdcall calc_crc32c_sw(uint32_t seed, uint8_t* msg, uint32_t msglen); */
PUBLIC calc_crc32c_sw
calc_crc32c_sw:
/* rax = crc / seed
* rdx = buf
* r8 = len
* rcx = tmp
* r10 = tmp2 */
mov rax, rcx
crcloop:
test r8, r8
jz crcend
mov rcx, rax
shr rcx, 8
mov r10b, byte ptr [rdx]
xor al, r10b
and rax, 255
shl rax, 2
lea r10, [rip+crctable]
mov eax, dword ptr [r10 + rax]
xor rax, rcx
inc rdx
dec r8
jmp crcloop
crcend:
ret
/****************************************************/
/* uint32_t __stdcall calc_crc32c_hw(uint32_t seed, uint8_t* msg, uint32_t msglen); */
PUBLIC calc_crc32c_hw
calc_crc32c_hw:
/* rax = crc / seed
* rdx = buf
* r8 = len */
mov rax, rcx
crchw_loop:
cmp r8, 8
jl crchw_stragglers
crc32 rax, qword ptr [rdx]
add rdx, 8
sub r8, 8
jmp crchw_loop
crchw_stragglers:
cmp r8, 4
jl crchw_stragglers2
crc32 eax, dword ptr [rdx]
add rdx, 4
sub r8, 4
crchw_stragglers2:
cmp r8, 2
jl crchw_stragglers3
crc32 eax, word ptr [rdx]
add rdx, 2
sub r8, 2
crchw_stragglers3:
test r8, r8
jz crchw_end
crc32 eax, byte ptr [rdx]
inc rdx
dec r8
jmp crchw_stragglers3
crchw_end:
ret
END
#elif defined(_X86_)
EXTERN _crctable:DWORD
.code
/* uint32_t __stdcall calc_crc32c_sw(uint32_t seed, uint8_t* msg, uint32_t msglen); */
@ -120,3 +217,4 @@ pop ebp
ret 12
END
#endif

8
drivers/filesystems/btrfs/crc32c.h
View file

@ -2,6 +2,10 @@
#include <stdint.h>
#ifdef __cplusplus
extern "C" {
#endif
#if defined(_X86_) || defined(_AMD64_)
uint32_t __stdcall calc_crc32c_hw(uint32_t seed, uint8_t* msg, uint32_t msglen);
#endif
@ -11,3 +15,7 @@ uint32_t __stdcall calc_crc32c_sw(uint32_t seed, uint8_t* msg, uint32_t msglen);
typedef uint32_t (__stdcall *crc_func)(uint32_t seed, uint8_t* msg, uint32_t msglen);
extern crc_func calc_crc32c;
#ifdef __cplusplus
}
#endif

8
drivers/filesystems/btrfs/create.c
View file

@ -387,7 +387,7 @@ static NTSTATUS split_path(device_extension* Vcb, PUNICODE_STRING path, LIST_ENT
if (nb->us.Buffer[i] == ':') {
name_bit* nb2;
if (nb->us.Buffer[i+1] == 0) {
if (i + 1 == nb->us.Length / sizeof(WCHAR)) {
WARN("zero-length stream name\n");
Status = STATUS_OBJECT_NAME_INVALID;
goto cleanup;
@ -709,7 +709,7 @@ NTSTATUS open_fcb(_Requires_lock_held_(_Curr_->tree_lock) _Requires_exclusive_lo
bool atts_set = false, sd_set = false, no_data;
LIST_ENTRY* lastle = NULL;
EXTENT_DATA* ed = NULL;
uint64_t fcbs_version;
uint64_t fcbs_version = 0;
uint32_t hash;
hash = calc_crc32c(0xffffffff, (uint8_t*)&inode, sizeof(uint64_t));
@ -1687,7 +1687,7 @@ NTSTATUS open_fileref(_Requires_lock_held_(_Curr_->tree_lock) _Requires_exclusiv
UNICODE_STRING fnus2;
file_ref *dir, *sf, *sf2;
LIST_ENTRY parts;
bool has_stream;
bool has_stream = false;
NTSTATUS Status;
LIST_ENTRY* le;
@ -4658,7 +4658,7 @@ loaded:
if (!NT_SUCCESS(Status))
goto exit;
} else {
file_ref* existing_file;
file_ref* existing_file = NULL;
Status = file_create(Irp, Vcb, FileObject, related, loaded_related, &fn, RequestedDisposition, options, &existing_file, rollback);

2
drivers/filesystems/btrfs/dirctrl.c
View file

@ -1174,7 +1174,7 @@ NTSTATUS __stdcall drv_directory_control(IN PDEVICE_OBJECT DeviceObject, IN PIRP
top_level = is_top_level(Irp);
if (Vcb && Vcb->type == VCB_TYPE_VOLUME) {
Status = vol_directory_control(DeviceObject, Irp);
Status = STATUS_INVALID_DEVICE_REQUEST;
goto end;
} else if (!Vcb || Vcb->type != VCB_TYPE_FS) {
Status = STATUS_INVALID_PARAMETER;

16
drivers/filesystems/btrfs/fastio.c
View file

@ -25,6 +25,8 @@ static BOOLEAN __stdcall fast_query_basic_info(PFILE_OBJECT FileObject, BOOLEAN
fcb* fcb;
ccb* ccb;
UNUSED(DeviceObject);
FsRtlEnterFileSystem();
TRACE("(%p, %u, %p, %p, %p)\n", FileObject, wait, fbi, IoStatus, DeviceObject);
@ -99,6 +101,8 @@ static BOOLEAN __stdcall fast_query_standard_info(PFILE_OBJECT FileObject, BOOLE
bool ads;
ULONG adssize;
UNUSED(DeviceObject);
FsRtlEnterFileSystem();
TRACE("(%p, %u, %p, %p, %p)\n", FileObject, wait, fsi, IoStatus, DeviceObject);
@ -198,6 +202,10 @@ static BOOLEAN __stdcall fast_io_query_network_open_info(PFILE_OBJECT FileObject
ccb* ccb;
file_ref* fileref;
UNUSED(Wait);
UNUSED(IoStatus); // FIXME - really? What about IoStatus->Information?
UNUSED(DeviceObject);
FsRtlEnterFileSystem();
TRACE("(%p, %u, %p, %p, %p)\n", FileObject, Wait, fnoi, IoStatus, DeviceObject);
@ -372,6 +380,8 @@ static BOOLEAN __stdcall fast_io_lock(PFILE_OBJECT FileObject, PLARGE_INTEGER Fi
BOOLEAN ret;
fcb* fcb = FileObject->FsContext;
UNUSED(DeviceObject);
TRACE("(%p, %I64x, %I64x, %p, %lx, %u, %u, %p, %p)\n", FileObject, FileOffset ? FileOffset->QuadPart : 0, Length ? Length->QuadPart : 0,
ProcessId, Key, FailImmediately, ExclusiveLock, IoStatus, DeviceObject);
@ -402,6 +412,8 @@ static BOOLEAN __stdcall fast_io_unlock_single(PFILE_OBJECT FileObject, PLARGE_I
ULONG Key, PIO_STATUS_BLOCK IoStatus, PDEVICE_OBJECT DeviceObject) {
fcb* fcb = FileObject->FsContext;
UNUSED(DeviceObject);
TRACE("(%p, %I64x, %I64x, %p, %lx, %p, %p)\n", FileObject, FileOffset ? FileOffset->QuadPart : 0, Length ? Length->QuadPart : 0,
ProcessId, Key, IoStatus, DeviceObject);
@ -428,6 +440,8 @@ _Function_class_(FAST_IO_UNLOCK_ALL)
static BOOLEAN __stdcall fast_io_unlock_all(PFILE_OBJECT FileObject, PEPROCESS ProcessId, PIO_STATUS_BLOCK IoStatus, PDEVICE_OBJECT DeviceObject) {
fcb* fcb = FileObject->FsContext;
UNUSED(DeviceObject);
TRACE("(%p, %p, %p, %p)\n", FileObject, ProcessId, IoStatus, DeviceObject);
IoStatus->Information = 0;
@ -458,6 +472,8 @@ static BOOLEAN __stdcall fast_io_unlock_all_by_key(PFILE_OBJECT FileObject, PVOI
PIO_STATUS_BLOCK IoStatus, PDEVICE_OBJECT DeviceObject) {
fcb* fcb = FileObject->FsContext;
UNUSED(DeviceObject);
TRACE("(%p, %p, %lx, %p, %p)\n", FileObject, ProcessId, Key, IoStatus, DeviceObject);
IoStatus->Information = 0;

8
drivers/filesystems/btrfs/fileinfo.c
View file

@ -3826,7 +3826,7 @@ NTSTATUS __stdcall drv_set_information(IN PDEVICE_OBJECT DeviceObject, IN PIRP I
Irp->IoStatus.Information = 0;
if (Vcb && Vcb->type == VCB_TYPE_VOLUME) {
Status = vol_set_information(DeviceObject, Irp);
Status = STATUS_INVALID_DEVICE_REQUEST;
goto end;
} else if (!Vcb || Vcb->type != VCB_TYPE_FS) {
Status = STATUS_INVALID_PARAMETER;
@ -5345,7 +5345,7 @@ NTSTATUS __stdcall drv_query_information(IN PDEVICE_OBJECT DeviceObject, IN PIRP
top_level = is_top_level(Irp);
if (Vcb && Vcb->type == VCB_TYPE_VOLUME) {
Status = vol_query_information(DeviceObject, Irp);
Status = STATUS_INVALID_DEVICE_REQUEST;
goto end;
} else if (!Vcb || Vcb->type != VCB_TYPE_FS) {
Status = STATUS_INVALID_PARAMETER;
@ -5399,7 +5399,7 @@ NTSTATUS __stdcall drv_query_ea(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
top_level = is_top_level(Irp);
if (Vcb && Vcb->type == VCB_TYPE_VOLUME) {
Status = vol_query_ea(DeviceObject, Irp);
Status = STATUS_INVALID_DEVICE_REQUEST;
goto end;
} else if (!Vcb || Vcb->type != VCB_TYPE_FS) {
Status = STATUS_INVALID_PARAMETER;
@ -5635,7 +5635,7 @@ NTSTATUS __stdcall drv_set_ea(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
top_level = is_top_level(Irp);
if (Vcb && Vcb->type == VCB_TYPE_VOLUME) {
Status = vol_set_ea(DeviceObject, Irp);
Status = STATUS_INVALID_DEVICE_REQUEST;
goto end;
} else if (!Vcb || Vcb->type != VCB_TYPE_FS) {
Status = STATUS_INVALID_PARAMETER;

5
drivers/filesystems/btrfs/flushthread.c
View file

@ -5816,6 +5816,8 @@ static NTSTATUS partial_stripe_read(device_extension* Vcb, chunk* c, partial_str
}
i = (parity + 1) % c->chunk_item->num_stripes;
logstripe = (c->chunk_item->num_stripes + c->chunk_item->num_stripes - 1 - parity + stripe) % c->chunk_item->num_stripes;
for (k = 0; k < c->chunk_item->num_stripes; k++) {
if (i != stripe) {
if (c->devices[i]->devobj) {
@ -5835,8 +5837,7 @@ static NTSTATUS partial_stripe_read(device_extension* Vcb, chunk* c, partial_str
ExFreePool(scratch);
return STATUS_UNEXPECTED_IO_ERROR;
}
} else
logstripe = k;
}
i = (i + 1) % c->chunk_item->num_stripes;
}

369
drivers/filesystems/btrfs/fsctl.c
View file

@ -17,6 +17,7 @@
#include "btrfs_drv.h"
#include "btrfsioctl.h"
#include "crc32c.h"
#include <ntddstor.h>
#include <ntdddisk.h>
#ifndef __REACTOS__
@ -2179,10 +2180,10 @@ end:
return Status;
}
static NTSTATUS get_object_id(device_extension* Vcb, PFILE_OBJECT FileObject, FILE_OBJECTID_BUFFER* buf, ULONG buflen, ULONG_PTR* retlen) {
static NTSTATUS get_object_id(PFILE_OBJECT FileObject, FILE_OBJECTID_BUFFER* buf, ULONG buflen, ULONG_PTR* retlen) {
fcb* fcb;
TRACE("(%p, %p, %p, %lx, %p)\n", Vcb, FileObject, buf, buflen, retlen);
TRACE("(%p, %p, %lx, %p)\n", FileObject, buf, buflen, retlen);
if (!FileObject) {
ERR("FileObject was NULL\n");
@ -4965,6 +4966,343 @@ static NTSTATUS fsctl_oplock(device_extension* Vcb, PIRP* Pirp) {
return Status;
}
static NTSTATUS get_retrieval_pointers(device_extension* Vcb, PFILE_OBJECT FileObject, STARTING_VCN_INPUT_BUFFER* in,
ULONG inlen, RETRIEVAL_POINTERS_BUFFER* out, ULONG outlen, ULONG_PTR* retlen) {
NTSTATUS Status;
fcb* fcb;
TRACE("get_retrieval_pointers(%p, %p, %p, %lx, %p, %lx, %p)\n", Vcb, FileObject, in, inlen,
out, outlen, retlen);
if (!FileObject)
return STATUS_INVALID_PARAMETER;
fcb = FileObject->FsContext;
if (!fcb)
return STATUS_INVALID_PARAMETER;
if (inlen < sizeof(STARTING_VCN_INPUT_BUFFER) || in->StartingVcn.QuadPart < 0)
return STATUS_INVALID_PARAMETER;
if (!out)
return STATUS_INVALID_PARAMETER;
if (outlen < offsetof(RETRIEVAL_POINTERS_BUFFER, Extents[0]))
return STATUS_BUFFER_TOO_SMALL;
ExAcquireResourceSharedLite(fcb->Header.Resource, true);
try {
LIST_ENTRY* le = fcb->extents.Flink;
extent* first_ext = NULL;
unsigned int num_extents = 0, first_extent_num = 0, i;
uint64_t num_sectors, last_off = 0;
num_sectors = (fcb->inode_item.st_size + Vcb->superblock.sector_size - 1) >> Vcb->sector_shift;
while (le != &fcb->extents) {
extent* ext = CONTAINING_RECORD(le, extent, list_entry);
if (ext->ignore || ext->extent_data.type == EXTENT_TYPE_INLINE) {
le = le->Flink;
continue;
}
if (ext->offset > last_off)
num_extents++;
if ((ext->offset >> Vcb->sector_shift) <= (uint64_t)in->StartingVcn.QuadPart &&
(ext->offset + ext->extent_data.decoded_size) >> Vcb->sector_shift > (uint64_t)in->StartingVcn.QuadPart) {
first_ext = ext;
first_extent_num = num_extents;
}
num_extents++;
last_off = ext->offset + ext->extent_data.decoded_size;
le = le->Flink;
}
if (num_sectors > last_off >> Vcb->sector_shift)
num_extents++;
if (!first_ext) {
Status = STATUS_END_OF_FILE;
leave;
}
out->ExtentCount = num_extents - first_extent_num;
out->StartingVcn.QuadPart = first_ext->offset >> Vcb->sector_shift;
outlen -= offsetof(RETRIEVAL_POINTERS_BUFFER, Extents[0]);
*retlen = offsetof(RETRIEVAL_POINTERS_BUFFER, Extents[0]);
le = &first_ext->list_entry;
i = 0;
last_off = 0;
while (le != &fcb->extents) {
extent* ext = CONTAINING_RECORD(le, extent, list_entry);
if (ext->ignore || ext->extent_data.type == EXTENT_TYPE_INLINE) {
le = le->Flink;
continue;
}
if (ext->offset > last_off) {
if (outlen < sizeof(LARGE_INTEGER) + sizeof(LARGE_INTEGER)) {
Status = STATUS_BUFFER_OVERFLOW;
leave;
}
out->Extents[i].NextVcn.QuadPart = ext->offset >> Vcb->sector_shift;
out->Extents[i].Lcn.QuadPart = -1;
outlen -= sizeof(LARGE_INTEGER) + sizeof(LARGE_INTEGER);
*retlen += sizeof(LARGE_INTEGER) + sizeof(LARGE_INTEGER);
i++;
}
if (outlen < sizeof(LARGE_INTEGER) + sizeof(LARGE_INTEGER)) {
Status = STATUS_BUFFER_OVERFLOW;
leave;
}
out->Extents[i].NextVcn.QuadPart = (ext->offset + ext->extent_data.decoded_size) >> Vcb->sector_shift;
if (ext->extent_data.compression == BTRFS_COMPRESSION_NONE) {
EXTENT_DATA2* ed2 = (EXTENT_DATA2*)ext->extent_data.data;
out->Extents[i].Lcn.QuadPart = (ed2->address + ed2->offset) >> Vcb->sector_shift;
} else
out->Extents[i].Lcn.QuadPart = -1;
outlen -= sizeof(LARGE_INTEGER) + sizeof(LARGE_INTEGER);
*retlen += sizeof(LARGE_INTEGER) + sizeof(LARGE_INTEGER);
i++;
le = le->Flink;
}
if (num_sectors << Vcb->sector_shift > last_off) {
if (outlen < sizeof(LARGE_INTEGER) + sizeof(LARGE_INTEGER)) {
Status = STATUS_BUFFER_OVERFLOW;
leave;
}
out->Extents[i].NextVcn.QuadPart = num_sectors;
out->Extents[i].Lcn.QuadPart = -1;
outlen -= sizeof(LARGE_INTEGER) + sizeof(LARGE_INTEGER);
*retlen += sizeof(LARGE_INTEGER) + sizeof(LARGE_INTEGER);
}
Status = STATUS_SUCCESS;
} finally {
ExReleaseResourceLite(fcb->Header.Resource);
}
return Status;
}
static NTSTATUS add_csum_sparse_extents(device_extension* Vcb, uint64_t sparse_extents, uint8_t** ptr, bool found, void* hash_ptr) {
if (!found) {
uint8_t* sector = ExAllocatePoolWithTag(PagedPool, Vcb->superblock.sector_size, ALLOC_TAG);
if (!sector) {
ERR("out of memory\n");
return STATUS_INSUFFICIENT_RESOURCES;
}
memset(sector, 0, Vcb->superblock.sector_size);
get_sector_csum(Vcb, sector, hash_ptr);
ExFreePool(sector);
}
switch (Vcb->superblock.csum_type) {
case CSUM_TYPE_CRC32C: {
uint32_t* csum = (uint32_t*)*ptr;
uint32_t sparse_hash = *(uint32_t*)hash_ptr;
for (uint64_t i = 0; i < sparse_extents; i++) {
csum[i] = sparse_hash;
}
break;
}
case CSUM_TYPE_XXHASH: {
uint64_t* csum = (uint64_t*)*ptr;
uint64_t sparse_hash = *(uint64_t*)hash_ptr;
for (uint64_t i = 0; i < sparse_extents; i++) {
csum[i] = sparse_hash;
}
break;
}
case CSUM_TYPE_SHA256:
case CSUM_TYPE_BLAKE2: {
uint8_t* csum = (uint8_t*)*ptr;
for (uint64_t i = 0; i < sparse_extents; i++) {
memcpy(csum, hash_ptr, 32);
csum += 32;
}
break;
}
default:
ERR("unrecognized hash type %x\n", Vcb->superblock.csum_type);
return STATUS_INTERNAL_ERROR;
}
*ptr += sparse_extents * Vcb->csum_size;
return STATUS_SUCCESS;
}
static NTSTATUS get_csum_info(device_extension* Vcb, PFILE_OBJECT FileObject, btrfs_csum_info* buf, ULONG buflen,
ULONG_PTR* retlen, KPROCESSOR_MODE processor_mode) {
NTSTATUS Status;
fcb* fcb;
ccb* ccb;
TRACE("get_csum_info(%p, %p, %p, %lx, %p, %x)\n", Vcb, FileObject, buf, buflen, retlen, processor_mode);
if (!FileObject)
return STATUS_INVALID_PARAMETER;
fcb = FileObject->FsContext;
ccb = FileObject->FsContext2;
if (!fcb || !ccb)
return STATUS_INVALID_PARAMETER;
if (!buf)
return STATUS_INVALID_PARAMETER;
if (buflen < offsetof(btrfs_csum_info, data[0]))
return STATUS_BUFFER_TOO_SMALL;
if (processor_mode == UserMode && !(ccb->access & (FILE_READ_DATA | FILE_WRITE_DATA))) {
WARN("insufficient privileges\n");
return STATUS_ACCESS_DENIED;
}
ExAcquireResourceSharedLite(fcb->Header.Resource, true);
try {
LIST_ENTRY* le;
uint8_t* ptr;
uint64_t last_off;
uint8_t sparse_hash[MAX_HASH_SIZE];
bool sparse_hash_found = false;
if (fcb->ads) {
Status = STATUS_INVALID_DEVICE_REQUEST;
leave;
}
if (fcb->type == BTRFS_TYPE_DIRECTORY) {
Status = STATUS_FILE_IS_A_DIRECTORY;
leave;
}
if (fcb->inode_item.flags & BTRFS_INODE_NODATASUM) {
Status = STATUS_INVALID_DEVICE_REQUEST;
leave;
}
buf->csum_type = Vcb->superblock.csum_type;
buf->csum_length = Vcb->csum_size;
le = fcb->extents.Flink;
while (le != &fcb->extents) {
extent* ext = CONTAINING_RECORD(le, extent, list_entry);
if (ext->ignore) {
le = le->Flink;
continue;
}
if (ext->extent_data.type == EXTENT_TYPE_INLINE) {
buf->num_sectors = 0;
*retlen = offsetof(btrfs_csum_info, data[0]);
Status = STATUS_SUCCESS;
leave;
}
le = le->Flink;
}
buf->num_sectors = (fcb->inode_item.st_size + Vcb->superblock.sector_size - 1) >> Vcb->sector_shift;
if (buflen < offsetof(btrfs_csum_info, data[0]) + (buf->csum_length * buf->num_sectors)) {
Status = STATUS_BUFFER_OVERFLOW;
*retlen = offsetof(btrfs_csum_info, data[0]);
leave;
}
ptr = buf->data;
last_off = 0;
le = fcb->extents.Flink;
while (le != &fcb->extents) {
extent* ext = CONTAINING_RECORD(le, extent, list_entry);
EXTENT_DATA2* ed2;
if (ext->ignore || ext->extent_data.type == EXTENT_TYPE_INLINE) {
le = le->Flink;
continue;
}
if (ext->offset > last_off) {
uint64_t sparse_extents = (ext->offset - last_off) >> Vcb->sector_shift;
add_csum_sparse_extents(Vcb, sparse_extents, &ptr, sparse_hash_found, sparse_hash);
sparse_hash_found = true;
}
ed2 = (EXTENT_DATA2*)ext->extent_data.data;
if (ext->extent_data.compression != BTRFS_COMPRESSION_NONE)
memset(ptr, 0, (ed2->num_bytes >> Vcb->sector_shift) * Vcb->csum_size); // dummy value for compressed extents
else {
if (ext->csum)
memcpy(ptr, ext->csum, (ed2->num_bytes >> Vcb->sector_shift) * Vcb->csum_size);
else
memset(ptr, 0, (ed2->num_bytes >> Vcb->sector_shift) * Vcb->csum_size);
ptr += (ed2->num_bytes >> Vcb->sector_shift) * Vcb->csum_size;
}
last_off = ext->offset + ed2->num_bytes;
le = le->Flink;
}
if (buf->num_sectors > last_off >> Vcb->sector_shift) {
uint64_t sparse_extents = buf->num_sectors - (last_off >> Vcb->sector_shift);
add_csum_sparse_extents(Vcb, sparse_extents, &ptr, sparse_hash_found, sparse_hash);
}
*retlen = offsetof(btrfs_csum_info, data[0]) + (buf->csum_length * buf->num_sectors);
Status = STATUS_SUCCESS;
} finally {
ExReleaseResourceLite(fcb->Header.Resource);
}
return Status;
}
NTSTATUS fsctl_request(PDEVICE_OBJECT DeviceObject, PIRP* Pirp, uint32_t type) {
PIRP Irp = *Pirp;
PIO_STACK_LOCATION IrpSp = IoGetCurrentIrpStackLocation(Irp);
@ -5065,8 +5403,11 @@ NTSTATUS fsctl_request(PDEVICE_OBJECT DeviceObject, PIRP* Pirp, uint32_t type) {
break;
case FSCTL_GET_RETRIEVAL_POINTERS:
WARN("STUB: FSCTL_GET_RETRIEVAL_POINTERS\n");
Status = STATUS_INVALID_DEVICE_REQUEST;
Status = get_retrieval_pointers(DeviceObject->DeviceExtension, IrpSp->FileObject,
IrpSp->Parameters.FileSystemControl.Type3InputBuffer,
IrpSp->Parameters.FileSystemControl.InputBufferLength,
Irp->UserBuffer, IrpSp->Parameters.FileSystemControl.OutputBufferLength,
&Irp->IoStatus.Information);
break;
case FSCTL_MOVE_FILE:
@ -5093,8 +5434,8 @@ NTSTATUS fsctl_request(PDEVICE_OBJECT DeviceObject, PIRP* Pirp, uint32_t type) {
break;
case FSCTL_GET_OBJECT_ID:
Status = get_object_id(DeviceObject->DeviceExtension, IrpSp->FileObject, Irp->UserBuffer,
IrpSp->Parameters.FileSystemControl.OutputBufferLength, &Irp->IoStatus.Information);
Status = get_object_id(IrpSp->FileObject, Irp->UserBuffer, IrpSp->Parameters.FileSystemControl.OutputBufferLength,
&Irp->IoStatus.Information);
break;
case FSCTL_DELETE_OBJECT_ID:
@ -5103,16 +5444,16 @@ NTSTATUS fsctl_request(PDEVICE_OBJECT DeviceObject, PIRP* Pirp, uint32_t type) {
break;
case FSCTL_SET_REPARSE_POINT:
Status = set_reparse_point(DeviceObject, Irp);
Status = set_reparse_point(Irp);
break;
case FSCTL_GET_REPARSE_POINT:
Status = get_reparse_point(DeviceObject, IrpSp->FileObject, Irp->AssociatedIrp.SystemBuffer,
Status = get_reparse_point(IrpSp->FileObject, Irp->AssociatedIrp.SystemBuffer,
IrpSp->Parameters.FileSystemControl.OutputBufferLength, &Irp->IoStatus.Information);
break;
case FSCTL_DELETE_REPARSE_POINT:
Status = delete_reparse_point(DeviceObject, Irp);
Status = delete_reparse_point(Irp);
break;
case FSCTL_ENUM_USN_DATA:
@ -5136,8 +5477,8 @@ NTSTATUS fsctl_request(PDEVICE_OBJECT DeviceObject, PIRP* Pirp, uint32_t type) {
break;
case FSCTL_CREATE_OR_GET_OBJECT_ID:
Status = get_object_id(DeviceObject->DeviceExtension, IrpSp->FileObject, Irp->UserBuffer,
IrpSp->Parameters.FileSystemControl.OutputBufferLength, &Irp->IoStatus.Information);
Status = get_object_id(IrpSp->FileObject, Irp->UserBuffer, IrpSp->Parameters.FileSystemControl.OutputBufferLength,
&Irp->IoStatus.Information);
break;
case FSCTL_SET_SPARSE:
@ -5558,6 +5899,12 @@ NTSTATUS fsctl_request(PDEVICE_OBJECT DeviceObject, PIRP* Pirp, uint32_t type) {
IrpSp->Parameters.FileSystemControl.InputBufferLength, Irp);
break;
case FSCTL_BTRFS_GET_CSUM_INFO:
Status = get_csum_info(DeviceObject->DeviceExtension, IrpSp->FileObject, Irp->AssociatedIrp.SystemBuffer,
IrpSp->Parameters.FileSystemControl.OutputBufferLength, &Irp->IoStatus.Information,
Irp->RequestorMode);
break;
default:
WARN("unknown control code %lx (DeviceType = %lx, Access = %lx, Function = %lx, Method = %lx)\n",
IrpSp->Parameters.FileSystemControl.FsControlCode, (IrpSp->Parameters.FileSystemControl.FsControlCode & 0xff0000) >> 16,

2
drivers/filesystems/btrfs/pnp.c
View file

@ -79,6 +79,8 @@ NTSTATUS pnp_surprise_removal(PDEVICE_OBJECT DeviceObject, PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
UNUSED(Irp);
if (DeviceObject->Vpb->Flags & VPB_MOUNTED) {
ExAcquireResourceExclusiveLite(&Vcb->tree_lock, true);

63
drivers/filesystems/btrfs/read.c
View file

@ -466,7 +466,7 @@ static NTSTATUS read_data_raid0(device_extension* Vcb, uint8_t* buf, uint64_t ad
static NTSTATUS read_data_raid10(device_extension* Vcb, uint8_t* buf, uint64_t addr, uint32_t length, read_data_context* context,
CHUNK_ITEM* ci, device** devices, uint64_t generation, uint64_t offset) {
uint16_t stripe;
uint16_t stripe = 0;
NTSTATUS Status;
bool checksum_error = false;
CHUNK_ITEM_STRIPE* cis = (CHUNK_ITEM_STRIPE*)&ci[1];
@ -653,7 +653,7 @@ static NTSTATUS read_data_raid5(device_extension* Vcb, uint8_t* buf, uint64_t ad
NTSTATUS Status;
bool checksum_error = false;
CHUNK_ITEM_STRIPE* cis = (CHUNK_ITEM_STRIPE*)&ci[1];
uint16_t j, stripe;
uint16_t j, stripe = 0;
bool no_success = true;
for (j = 0; j < ci->num_stripes; j++) {
@ -950,7 +950,7 @@ void raid6_recover2(uint8_t* sectors, uint16_t num_stripes, ULONG sector_size, u
if (missing != 0)
galois_divpower(out, (uint8_t)missing, sector_size);
} else { // reconstruct from p and q
uint16_t x, y, stripe;
uint16_t x = missing1, y = missing2, stripe;
uint8_t gyx, gx, denom, a, b, *p, *q, *pxy, *qxy;
uint32_t j;
@ -962,11 +962,6 @@ void raid6_recover2(uint8_t* sectors, uint16_t num_stripes, ULONG sector_size, u
if (stripe == missing1 || stripe == missing2) {
RtlZeroMemory(qxy, sector_size);
RtlZeroMemory(pxy, sector_size);
if (stripe == missing1)
x = stripe;
else
y = stripe;
} else {
RtlCopyMemory(qxy, sectors + (stripe * sector_size), sector_size);
RtlCopyMemory(pxy, sectors + (stripe * sector_size), sector_size);
@ -980,10 +975,7 @@ void raid6_recover2(uint8_t* sectors, uint16_t num_stripes, ULONG sector_size, u
if (stripe != missing1 && stripe != missing2) {
do_xor(qxy, sectors + (stripe * sector_size), sector_size);
do_xor(pxy, sectors + (stripe * sector_size), sector_size);
} else if (stripe == missing1)
x = stripe;
else if (stripe == missing2)
y = stripe;
}
} while (stripe > 0);
gyx = gpow2(y > x ? (y-x) : (255-x+y));
@ -1015,7 +1007,7 @@ static NTSTATUS read_data_raid6(device_extension* Vcb, uint8_t* buf, uint64_t ad
NTSTATUS Status;
bool checksum_error = false;
CHUNK_ITEM_STRIPE* cis = (CHUNK_ITEM_STRIPE*)&ci[1];
uint16_t stripe, j;
uint16_t stripe = 0, j;
bool no_success = true;
for (j = 0; j < ci->num_stripes; j++) {
@ -1117,7 +1109,7 @@ static NTSTATUS read_data_raid6(device_extension* Vcb, uint8_t* buf, uint64_t ad
if (context->tree) {
uint8_t* sector;
uint16_t k, physstripe, parity1, parity2, error_stripe;
uint16_t k, physstripe, parity1, parity2, error_stripe = 0;
uint64_t off;
bool recovered = false, failed = false;
ULONG num_errors = 0;
@ -1321,7 +1313,7 @@ static NTSTATUS read_data_raid6(device_extension* Vcb, uint8_t* buf, uint64_t ad
physstripe = (parity2 + stripe + 1) % ci->num_stripes;
if (!devices[physstripe] || !devices[physstripe]->devobj || (context->csum && !check_sector_csum(Vcb, buf + (i << Vcb->sector_shift), ptr))) {
uint16_t error_stripe;
uint16_t error_stripe = 0;
bool recovered = false, failed = false;
ULONG num_errors = 0;
@ -2018,7 +2010,7 @@ NTSTATUS read_data(_In_ device_extension* Vcb, _In_ uint64_t addr, _In_ uint32_t
uint16_t endoffstripe, parity;
uint32_t *stripeoff, pos;
PMDL master_mdl;
PFN_NUMBER *pfns, dummy;
PFN_NUMBER *pfns, dummy = 0;
bool need_dummy = false;
get_raid0_offset(addr - offset, ci->stripe_length, ci->num_stripes - 1, &startoff, &startoffstripe);
@ -2277,7 +2269,7 @@ NTSTATUS read_data(_In_ device_extension* Vcb, _In_ uint64_t addr, _In_ uint32_t
uint16_t endoffstripe, parity1;
uint32_t *stripeoff, pos;
PMDL master_mdl;
PFN_NUMBER *pfns, dummy;
PFN_NUMBER *pfns, dummy = 0;
bool need_dummy = false;
get_raid0_offset(addr - offset, ci->stripe_length, ci->num_stripes - 2, &startoff, &startoffstripe);
@ -3243,8 +3235,14 @@ nextitem:
inpageoff = inoff % LZO_PAGE_SIZE;
}
if (off2 != 0) {
/* Previous versions of this code decompressed directly into the destination buffer,
* but unfortunately that can't be relied on - Windows likes to use dummy pages sometimes
* when mmap-ing, which breaks the backtracking used by e.g. zstd. */
if (off2 != 0)
outlen = off2 + min(rp->read, (uint32_t)(rp->extents[i].ed_num_bytes - rp->extents[i].off));
else
outlen = min(rp->read, (uint32_t)(rp->extents[i].ed_num_bytes - rp->extents[i].off));
decomp = ExAllocatePoolWithTag(pool_type, outlen, ALLOC_TAG);
if (!decomp) {
@ -3252,39 +3250,34 @@ nextitem:
Status = STATUS_INSUFFICIENT_RESOURCES;
goto exit;
}
} else
outlen = min(rp->read, (uint32_t)(rp->extents[i].ed_num_bytes - rp->extents[i].off));
ccj = (comp_calc_job*)ExAllocatePoolWithTag(pool_type, sizeof(comp_calc_job), ALLOC_TAG);
if (!ccj) {
ERR("out of memory\n");
if (decomp)
ExFreePool(decomp);
Status = STATUS_INSUFFICIENT_RESOURCES;
goto exit;
}
Status = add_calc_job_decomp(fcb->Vcb, rp->compression, buf2, inlen, decomp ? decomp : rp->data, outlen,
inpageoff, &ccj->cj);
if (!NT_SUCCESS(Status)) {
ERR("add_calc_job_decomp returned %08lx\n", Status);
if (decomp)
ExFreePool(decomp);
ExFreePool(ccj);
goto exit;
}
ccj->data = rp->data;
ccj->decomp = decomp;
ccj->offset = off2;
ccj->length = (size_t)min(rp->read, rp->extents[i].ed_num_bytes - rp->extents[i].off);
Status = add_calc_job_decomp(fcb->Vcb, rp->compression, buf2, inlen, decomp, outlen,
inpageoff, &ccj->cj);
if (!NT_SUCCESS(Status)) {
ERR("add_calc_job_decomp returned %08lx\n", Status);
ExFreePool(decomp);
ExFreePool(ccj);
goto exit;
}
InsertTailList(&calc_jobs, &ccj->list_entry);
buf += rp->extents[i].ed_size;
@ -3317,10 +3310,8 @@ nextitem:
if (!NT_SUCCESS(ccj->cj->Status))
Status = ccj->cj->Status;
if (ccj->decomp) {
RtlCopyMemory(ccj->data, (uint8_t*)ccj->decomp + ccj->offset, ccj->length);
ExFreePool(ccj->decomp);
}
ExFreePool(ccj);
}

19
drivers/filesystems/btrfs/reparse.c
View file

@ -24,7 +24,7 @@ typedef struct {
char name[1];
} REPARSE_DATA_BUFFER_LX_SYMLINK;
NTSTATUS get_reparse_point(PDEVICE_OBJECT DeviceObject, PFILE_OBJECT FileObject, void* buffer, DWORD buflen, ULONG_PTR* retlen) {
NTSTATUS get_reparse_point(PFILE_OBJECT FileObject, void* buffer, DWORD buflen, ULONG_PTR* retlen) {
USHORT subnamelen, printnamelen, i;
ULONG stringlen;
DWORD reqlen;
@ -33,7 +33,7 @@ NTSTATUS get_reparse_point(PDEVICE_OBJECT DeviceObject, PFILE_OBJECT FileObject,
ccb* ccb = FileObject->FsContext2;
NTSTATUS Status;
TRACE("(%p, %p, %p, %lx, %p)\n", DeviceObject, FileObject, buffer, buflen, retlen);
TRACE("(%p, %p, %lx, %p)\n", FileObject, buffer, buflen, retlen);
if (!ccb)
return STATUS_INVALID_PARAMETER;
@ -234,15 +234,16 @@ static NTSTATUS set_symlink(PIRP Irp, file_ref* fileref, fcb* fcb, ccb* ccb, REP
REPARSE_DATA_BUFFER_LX_SYMLINK* buf;
if (buflen < offsetof(REPARSE_DATA_BUFFER, GenericReparseBuffer.DataBuffer) + rdb->ReparseDataLength) {
WARN("buffer was less than expected length (%lu < %u)\n", buflen,
offsetof(REPARSE_DATA_BUFFER, GenericReparseBuffer.DataBuffer) + rdb->ReparseDataLength);
WARN("buffer was less than expected length (%lu < %lu)\n", buflen,
(unsigned long)(offsetof(REPARSE_DATA_BUFFER, GenericReparseBuffer.DataBuffer) + rdb->ReparseDataLength));
return STATUS_INVALID_PARAMETER;
}
buf = (REPARSE_DATA_BUFFER_LX_SYMLINK*)rdb->GenericReparseBuffer.DataBuffer;
if (buflen < offsetof(REPARSE_DATA_BUFFER_LX_SYMLINK, name)) {
WARN("buffer was less than minimum length (%u < %u)\n", rdb->ReparseDataLength, offsetof(REPARSE_DATA_BUFFER_LX_SYMLINK, name));
WARN("buffer was less than minimum length (%u < %lu)\n", rdb->ReparseDataLength,
(unsigned long)(offsetof(REPARSE_DATA_BUFFER_LX_SYMLINK, name)));
return STATUS_INVALID_PARAMETER;
}
@ -397,7 +398,7 @@ NTSTATUS set_reparse_point2(fcb* fcb, REPARSE_DATA_BUFFER* rdb, ULONG buflen, cc
return STATUS_SUCCESS;
}
NTSTATUS set_reparse_point(PDEVICE_OBJECT DeviceObject, PIRP Irp) {
NTSTATUS set_reparse_point(PIRP Irp) {
PIO_STACK_LOCATION IrpSp = IoGetCurrentIrpStackLocation(Irp);
PFILE_OBJECT FileObject = IrpSp->FileObject;
void* buffer = Irp->AssociatedIrp.SystemBuffer;
@ -409,7 +410,7 @@ NTSTATUS set_reparse_point(PDEVICE_OBJECT DeviceObject, PIRP Irp) {
file_ref* fileref;
LIST_ENTRY rollback;
TRACE("(%p, %p)\n", DeviceObject, Irp);
TRACE("(%p)\n", Irp);
InitializeListHead(&rollback);
@ -470,7 +471,7 @@ end:
return Status;
}
NTSTATUS delete_reparse_point(PDEVICE_OBJECT DeviceObject, PIRP Irp) {
NTSTATUS delete_reparse_point(PIRP Irp) {
PIO_STACK_LOCATION IrpSp = IoGetCurrentIrpStackLocation(Irp);
PFILE_OBJECT FileObject = IrpSp->FileObject;
REPARSE_DATA_BUFFER* rdb = Irp->AssociatedIrp.SystemBuffer;
@ -481,7 +482,7 @@ NTSTATUS delete_reparse_point(PDEVICE_OBJECT DeviceObject, PIRP Irp) {
file_ref* fileref;
LIST_ENTRY rollback;
TRACE("(%p, %p)\n", DeviceObject, Irp);
TRACE("(%p)\n", Irp);
InitializeListHead(&rollback);

19
drivers/filesystems/btrfs/scrub.c
View file

@ -1377,7 +1377,7 @@ static NTSTATUS scrub_extent(device_extension* Vcb, chunk* c, ULONG type, uint64
scrub_context context;
CHUNK_ITEM_STRIPE* cis;
NTSTATUS Status;
uint16_t startoffstripe, num_missing, allowed_missing;
uint16_t startoffstripe = 0, num_missing, allowed_missing;
TRACE("(%p, %p, %lx, %I64x, %x, %p)\n", Vcb, c, type, offset, size, csum);
@ -1788,8 +1788,8 @@ static void scrub_raid5_stripe(device_extension* Vcb, chunk* c, scrub_context_ra
return;
for (ULONG i = 0; i < sectors_per_stripe; i++) {
ULONG num_errors = 0, bad_off;
uint64_t bad_stripe;
ULONG num_errors = 0, bad_off = 0;
uint64_t bad_stripe = 0;
bool alloc = false;
stripe = (parity + 1) % c->chunk_item->num_stripes;
@ -2006,8 +2006,8 @@ static void scrub_raid6_stripe(device_extension* Vcb, chunk* c, scrub_context_ra
for (ULONG i = 0; i < sectors_per_stripe; i++) {
ULONG num_errors = 0;
uint64_t bad_stripe1, bad_stripe2;
ULONG bad_off1, bad_off2;
uint64_t bad_stripe1 = 0, bad_stripe2 = 0;
ULONG bad_off1 = 0, bad_off2 = 0;
bool alloc = false;
stripe = (parity1 + 2) % c->chunk_item->num_stripes;
@ -2071,7 +2071,6 @@ static void scrub_raid6_stripe(device_extension* Vcb, chunk* c, scrub_context_ra
}
} else if (num_errors == 1) {
uint32_t len;
uint16_t stripe_num, bad_stripe_num;
uint64_t addr = c->offset + (stripe_start * (c->chunk_item->num_stripes - 2) * c->chunk_item->stripe_length) + (bad_off1 << Vcb->sector_shift);
uint8_t* scratch;
@ -2091,6 +2090,8 @@ static void scrub_raid6_stripe(device_extension* Vcb, chunk* c, scrub_context_ra
stripe = parity1 == 0 ? (c->chunk_item->num_stripes - 1) : (parity1 - 1);
if (c->devices[parity2]->devobj) {
uint16_t stripe_num, bad_stripe_num = 0;
stripe_num = c->chunk_item->num_stripes - 3;
while (stripe != parity2) {
galois_double(scratch, len);
@ -2275,7 +2276,7 @@ static void scrub_raid6_stripe(device_extension* Vcb, chunk* c, scrub_context_ra
ExFreePool(scratch);
} else if (num_errors == 2 && missing_devices == 0) {
uint16_t x, y, k;
uint16_t x = 0, y = 0, k;
uint64_t addr;
uint32_t len = (RtlCheckBit(&context->is_tree, bad_off1) || RtlCheckBit(&context->is_tree, bad_off2)) ? Vcb->superblock.node_size : Vcb->superblock.sector_size;
uint8_t gyx, gx, denom, a, b, *p, *q, *pxy, *qxy;
@ -2837,7 +2838,7 @@ static NTSTATUS scrub_chunk_raid56(device_extension* Vcb, chunk* c, uint64_t* of
KEY searchkey;
traverse_ptr tp;
bool b;
uint64_t full_stripe_len, stripe, stripe_start, stripe_end, total_data = 0;
uint64_t full_stripe_len, stripe, stripe_start = 0, stripe_end = 0, total_data = 0;
ULONG num_extents = 0, num_parity_stripes = c->chunk_item->type & BLOCK_FLAG_RAID6 ? 2 : 1;
full_stripe_len = (c->chunk_item->num_stripes - num_parity_stripes) * c->chunk_item->stripe_length;
@ -2925,7 +2926,7 @@ static NTSTATUS scrub_chunk(device_extension* Vcb, chunk* c, uint64_t* offset, b
traverse_ptr tp;
bool b = false, tree_run = false;
ULONG type, num_extents = 0;
uint64_t total_data = 0, tree_run_start, tree_run_end;
uint64_t total_data = 0, tree_run_start = 0, tree_run_end = 0;
TRACE("chunk %I64x\n", c->offset);

4
drivers/filesystems/btrfs/security.c
View file

@ -678,7 +678,7 @@ NTSTATUS __stdcall drv_query_security(IN PDEVICE_OBJECT DeviceObject, IN PIRP Ir
top_level = is_top_level(Irp);
if (Vcb && Vcb->type == VCB_TYPE_VOLUME) {
Status = vol_query_security(DeviceObject, Irp);
Status = STATUS_INVALID_DEVICE_REQUEST;
goto end;
} else if (!Vcb || Vcb->type != VCB_TYPE_FS) {
Status = STATUS_INVALID_PARAMETER;
@ -842,7 +842,7 @@ NTSTATUS __stdcall drv_set_security(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp)
top_level = is_top_level(Irp);
if (Vcb && Vcb->type == VCB_TYPE_VOLUME) {
Status = vol_set_security(DeviceObject, Irp);
Status = STATUS_INVALID_DEVICE_REQUEST;
goto end;
} else if (!Vcb || Vcb->type != VCB_TYPE_FS) {
Status = STATUS_INVALID_PARAMETER;

94
drivers/filesystems/btrfs/volume.c
View file

@ -303,74 +303,6 @@ end:
return Status;
}
NTSTATUS vol_query_information(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
return STATUS_INVALID_DEVICE_REQUEST;
}
NTSTATUS vol_set_information(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
return STATUS_INVALID_DEVICE_REQUEST;
}
NTSTATUS vol_query_ea(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
return STATUS_INVALID_DEVICE_REQUEST;
}
NTSTATUS vol_set_ea(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
return STATUS_INVALID_DEVICE_REQUEST;
}
NTSTATUS vol_flush_buffers(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
return STATUS_SUCCESS;
}
NTSTATUS vol_query_volume_information(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
return STATUS_INVALID_DEVICE_REQUEST;
}
NTSTATUS vol_set_volume_information(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
return STATUS_INVALID_DEVICE_REQUEST;
}
NTSTATUS vol_cleanup(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
Irp->IoStatus.Information = 0;
return STATUS_SUCCESS;
}
NTSTATUS vol_directory_control(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
return STATUS_INVALID_DEVICE_REQUEST;
}
NTSTATUS vol_file_system_control(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
return STATUS_INVALID_DEVICE_REQUEST;
}
NTSTATUS vol_lock_control(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
return STATUS_INVALID_DEVICE_REQUEST;
}
static NTSTATUS vol_query_device_name(volume_device_extension* vde, PIRP Irp) {
PIO_STACK_LOCATION IrpSp = IoGetCurrentIrpStackLocation(Irp);
PMOUNTDEV_NAME name;
@ -877,17 +809,15 @@ NTSTATUS vol_device_control(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
return vol_get_disk_extents(vde, Irp);
default: { // pass ioctl through if only one child device
ULONG code = IrpSp->Parameters.DeviceIoControl.IoControlCode;
NTSTATUS Status = vol_ioctl_passthrough(vde, Irp);
#ifdef __REACTOS__
&code;
#endif
#ifdef _DEBUG
ULONG code = IrpSp->Parameters.DeviceIoControl.IoControlCode;
if (NT_SUCCESS(Status))
TRACE("passing through ioctl %lx (returning %08lx)\n", code, Status);
else
WARN("passing through ioctl %lx (returning %08lx)\n", code, Status);
#endif
return Status;
}
@ -896,24 +826,6 @@ NTSTATUS vol_device_control(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
return STATUS_INVALID_DEVICE_REQUEST;
}
NTSTATUS vol_shutdown(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
return STATUS_INVALID_DEVICE_REQUEST;
}
NTSTATUS vol_query_security(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
return STATUS_INVALID_DEVICE_REQUEST;
}
NTSTATUS vol_set_security(IN PDEVICE_OBJECT DeviceObject, IN PIRP Irp) {
TRACE("(%p, %p)\n", DeviceObject, Irp);
return STATUS_INVALID_DEVICE_REQUEST;
}
NTSTATUS mountmgr_add_drive_letter(PDEVICE_OBJECT mountmgr, PUNICODE_STRING devpath) {
NTSTATUS Status;
ULONG mmdltsize;

6
drivers/filesystems/btrfs/write.c
View file

@ -552,9 +552,7 @@ NTSTATUS alloc_chunk(device_extension* Vcb, uint64_t flags, chunk** pc, bool ful
}
}
if (type == 0 || type == BLOCK_FLAG_DUPLICATE || type == BLOCK_FLAG_RAID1 || type == BLOCK_FLAG_RAID1C3 || type == BLOCK_FLAG_RAID1C4)
factor = 1;
else if (type == BLOCK_FLAG_RAID0)
if (type == BLOCK_FLAG_RAID0)
factor = num_stripes;
else if (type == BLOCK_FLAG_RAID10)
factor = num_stripes / sub_stripes;
@ -562,6 +560,8 @@ NTSTATUS alloc_chunk(device_extension* Vcb, uint64_t flags, chunk** pc, bool ful
factor = num_stripes - 1;
else if (type == BLOCK_FLAG_RAID6)
factor = num_stripes - 2;
else
factor = 1; // SINGLE, DUPLICATE, RAID1, RAID1C3, RAID1C4
if (stripe_size * factor > max_chunk_size)
stripe_size = max_chunk_size / factor;

342
drivers/filesystems/btrfs/xor.S Normal file
View file

@ -0,0 +1,342 @@
/* Copyright (c) Mark Harmstone 2020
*
* This file is part of WinBtrfs.
*
* WinBtrfs is free software: you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public Licence as published by
* the Free Software Foundation, either version 3 of the Licence, or
* (at your option) any later version.
*
* WinBtrfs is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Lesser General Public Licence for more details.
*
* You should have received a copy of the GNU Lesser General Public Licence
* along with WinBtrfs. If not, see <http://www.gnu.org/licenses/>. */
#include <asm.inc>
#ifdef __x86_64__
.code64
/* void do_xor_sse2(uint8_t* buf1, uint8_t* buf2, uint32_t len); */
PUBLIC do_xor_sse2
do_xor_sse2:
/* rcx = buf1
* rdx = buf2
* r8d = len
* rax = tmp1
* r9 = tmp2
* xmm0 = tmp3
* xmm1 = tmp4 */
mov rax, rcx
and rax, 15
cmp rax, 0
jne stragglers2
mov rax, rdx
and rax, 15
cmp rax, 0
jne stragglers2
do_xor_sse2_loop:
cmp r8d, 16
jl stragglers2
movdqa xmm0, [rcx]
movdqa xmm1, [rdx]
pxor xmm0, xmm1
movdqa [rcx], xmm0
add rcx, 16
add rdx, 16
sub r8d, 16
jmp do_xor_sse2_loop
stragglers2:
cmp r8d, 8
jl stragglers
mov rax, [rcx]
mov r9, [rdx]
xor rax, r9
mov [rcx], rax
add rcx, 8
add rdx, 8
sub r8d, 8
jmp stragglers2
stragglers:
cmp r8d, 0
je do_xor_sse2_end
mov al, [rcx]
mov r9b, [rdx]
xor al, r9b
mov [rcx], al
inc rcx
inc rdx
dec r8d
jmp stragglers
do_xor_sse2_end:
ret
/* void do_xor_avx2(uint8_t* buf1, uint8_t* buf2, uint32_t len); */
PUBLIC do_xor_avx2
do_xor_avx2:
/* rcx = buf1
* rdx = buf2
* r8d = len
* rax = tmp1
* r9 = tmp2
* xmm0 = tmp3
* xmm1 = tmp4 */
mov rax, rcx
and rax, 31
cmp rax, 0
jne stragglers4
mov rax, rdx
and rax, 31
cmp rax, 0
jne stragglers4
do_xor_avx2_loop:
cmp r8d, 32
jl stragglers4
vmovdqa ymm0, YMMWORD PTR[rcx]
vmovdqa ymm1, YMMWORD PTR[rdx]
vpxor ymm0, ymm0, ymm1
vmovdqa YMMWORD PTR[rcx], ymm0
add rcx, 32
add rdx, 32
sub r8d, 32
jmp do_xor_avx2_loop
stragglers4:
cmp r8d, 8
jl stragglers3
mov rax, [rcx]
mov r9, [rdx]
xor rax, r9
mov [rcx], rax
add rcx, 8
add rdx, 8
sub r8d, 8
jmp stragglers4
stragglers3:
cmp r8d, 0
je do_xor_avx2_end
mov al, [rcx]
mov r9b, [rdx]
xor al, r9b
mov [rcx], al
inc rcx
inc rdx
dec r8d
jmp stragglers3
do_xor_avx2_end:
ret
END
#else
.code
/* void __stdcall do_xor_sse2(uint8_t* buf1, uint8_t* buf2, uint32_t len); */
PUBLIC _do_xor_sse2@12
_do_xor_sse2@12:
/* edi = buf1
* edx = buf2
* esi = len
* eax = tmp1
* ecx = tmp2
* xmm0 = tmp3
* xmm1 = tmp4 */
push ebp
mov ebp, esp
push esi
push edi
mov edi, [ebp+8]
mov edx, [ebp+12]
mov esi, [ebp+16]
mov eax, edi
and eax, 15
cmp eax, 0
jne stragglers2
mov eax, edx
and eax, 15
cmp eax, 0
jne stragglers2
do_xor_sse2_loop:
cmp esi, 16
jl stragglers2
movdqa xmm0, [edi]
movdqa xmm1, [edx]
pxor xmm0, xmm1
movdqa [edi], xmm0
add edi, 16
add edx, 16
sub esi, 16
jmp do_xor_sse2_loop
stragglers2:
cmp esi, 4
jl stragglers
mov eax, [edi]
mov ecx, [edx]
xor eax, ecx
mov [edi], eax
add edi, 4
add edx, 4
sub esi, 4
jmp stragglers2
stragglers:
cmp esi, 0
je do_xor_sse2_end
mov al, [edi]
mov cl, [edx]
xor al, cl
mov [edi], al
inc edi
inc edx
dec esi
jmp stragglers
do_xor_sse2_end:
pop edi
pop esi
pop ebp
ret 12
/* void __stdcall do_xor_avx2(uint8_t* buf1, uint8_t* buf2, uint32_t len); */
PUBLIC _do_xor_avx2@12
_do_xor_avx2@12:
/* edi = buf1
* edx = buf2
* esi = len
* eax = tmp1
* ecx = tmp2
* xmm0 = tmp3
* xmm1 = tmp4 */
push ebp
mov ebp, esp
push esi
push edi
mov edi, [ebp+8]
mov edx, [ebp+12]
mov esi, [ebp+16]
mov eax, edi
and eax, 31
cmp eax, 0
jne stragglers4
mov eax, edx
and eax, 31
cmp eax, 0
jne stragglers4
do_xor_avx2_loop:
cmp esi, 32
jl stragglers4
vmovdqa ymm0, YMMWORD PTR[edi]
vmovdqa ymm1, YMMWORD PTR[edx]
vpxor ymm0, ymm0, ymm1
vmovdqa YMMWORD PTR[edi], ymm0
add edi, 32
add edx, 32
sub esi, 32
jmp do_xor_avx2_loop
stragglers4:
cmp esi, 4
jl stragglers3
mov eax, [edi]
mov ecx, [edx]
xor eax, ecx
mov [edi], eax
add edi, 4
add edx, 4
sub esi, 4
jmp stragglers4
stragglers3:
cmp esi, 0
je do_xor_avx2_end
mov al, [edi]
mov cl, [edx]
xor al, cl
mov [edi], al
inc edi
inc edx
dec esi
jmp stragglers3
do_xor_avx2_end:
pop edi
pop esi
pop ebp
ret 12
END
#endif

81
drivers/filesystems/btrfs/zstd/bitstream.h
View file

@ -1,35 +1,15 @@
/* ******************************************************************
bitstream
Part of FSE library
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- Source repository : https://github.com/Cyan4973/FiniteStateEntropy
* bitstream
* Part of FSE library
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - Source repository : https://github.com/Cyan4973/FiniteStateEntropy
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
#ifndef BITSTREAM_H_MODULE
#define BITSTREAM_H_MODULE
@ -48,6 +28,7 @@ extern "C" {
* Dependencies
******************************************/
#include "mem.h" /* unaligned access routines */
#include "compiler.h" /* UNLIKELY() */
#include "debug.h" /* assert(), DEBUGLOG(), RAWLOG() */
#include "error_private.h" /* error codes and messages */
@ -57,6 +38,8 @@ extern "C" {
=========================================*/
#if defined(__BMI__) && defined(__GNUC__)
# include <immintrin.h> /* support for bextr (experimental) */
#elif defined(__ICCARM__)
# include <intrinsics.h>
#endif
#define STREAM_ACCUMULATOR_MIN_32 25
@ -159,10 +142,11 @@ MEM_STATIC unsigned BIT_highbit32 (U32 val)
{
# if defined(_MSC_VER) /* Visual */
unsigned long r=0;
_BitScanReverse ( &r, val );
return (unsigned) r;
return _BitScanReverse ( &r, val ) ? (unsigned)r : 0;
# elif defined(__GNUC__) && (__GNUC__ >= 3) /* Use GCC Intrinsic */
return 31 - __builtin_clz (val);
return __builtin_clz (val) ^ 31;
# elif defined(__ICCARM__) /* IAR Intrinsic */
return 31 - __CLZ(val);
# else /* Software version */
static const unsigned DeBruijnClz[32] = { 0, 9, 1, 10, 13, 21, 2, 29,
11, 14, 16, 18, 22, 25, 3, 30,
@ -240,9 +224,9 @@ MEM_STATIC void BIT_flushBitsFast(BIT_CStream_t* bitC)
{
size_t const nbBytes = bitC->bitPos >> 3;
assert(bitC->bitPos < sizeof(bitC->bitContainer) * 8);
assert(bitC->ptr <= bitC->endPtr);
MEM_writeLEST(bitC->ptr, bitC->bitContainer);
bitC->ptr += nbBytes;
assert(bitC->ptr <= bitC->endPtr);
bitC->bitPos &= 7;
bitC->bitContainer >>= nbBytes*8;
}
@ -256,6 +240,7 @@ MEM_STATIC void BIT_flushBits(BIT_CStream_t* bitC)
{
size_t const nbBytes = bitC->bitPos >> 3;
assert(bitC->bitPos < sizeof(bitC->bitContainer) * 8);
assert(bitC->ptr <= bitC->endPtr);
MEM_writeLEST(bitC->ptr, bitC->bitContainer);
bitC->ptr += nbBytes;
if (bitC->ptr > bitC->endPtr) bitC->ptr = bitC->endPtr;
@ -389,7 +374,7 @@ MEM_STATIC void BIT_skipBits(BIT_DStream_t* bitD, U32 nbBits)
* Read (consume) next n bits from local register and update.
* Pay attention to not read more than nbBits contained into local register.
* @return : extracted value. */
MEM_STATIC size_t BIT_readBits(BIT_DStream_t* bitD, U32 nbBits)
MEM_STATIC size_t BIT_readBits(BIT_DStream_t* bitD, unsigned nbBits)
{
size_t const value = BIT_lookBits(bitD, nbBits);
BIT_skipBits(bitD, nbBits);
@ -398,7 +383,7 @@ MEM_STATIC size_t BIT_readBits(BIT_DStream_t* bitD, U32 nbBits)
/*! BIT_readBitsFast() :
* unsafe version; only works only if nbBits >= 1 */
MEM_STATIC size_t BIT_readBitsFast(BIT_DStream_t* bitD, U32 nbBits)
MEM_STATIC size_t BIT_readBitsFast(BIT_DStream_t* bitD, unsigned nbBits)
{
size_t const value = BIT_lookBitsFast(bitD, nbBits);
assert(nbBits >= 1);
@ -406,6 +391,23 @@ MEM_STATIC size_t BIT_readBitsFast(BIT_DStream_t* bitD, U32 nbBits)
return value;
}
/*! BIT_reloadDStreamFast() :
* Similar to BIT_reloadDStream(), but with two differences:
* 1. bitsConsumed <= sizeof(bitD->bitContainer)*8 must hold!
* 2. Returns BIT_DStream_overflow when bitD->ptr < bitD->limitPtr, at this
* point you must use BIT_reloadDStream() to reload.
*/
MEM_STATIC BIT_DStream_status BIT_reloadDStreamFast(BIT_DStream_t* bitD)
{
if (UNLIKELY(bitD->ptr < bitD->limitPtr))
return BIT_DStream_overflow;
assert(bitD->bitsConsumed <= sizeof(bitD->bitContainer)*8);
bitD->ptr -= bitD->bitsConsumed >> 3;
bitD->bitsConsumed &= 7;
bitD->bitContainer = MEM_readLEST(bitD->ptr);
return BIT_DStream_unfinished;
}
/*! BIT_reloadDStream() :
* Refill `bitD` from buffer previously set in BIT_initDStream() .
* This function is safe, it guarantees it will not read beyond src buffer.
@ -417,10 +419,7 @@ MEM_STATIC BIT_DStream_status BIT_reloadDStream(BIT_DStream_t* bitD)
return BIT_DStream_overflow;
if (bitD->ptr >= bitD->limitPtr) {
bitD->ptr -= bitD->bitsConsumed >> 3;
bitD->bitsConsumed &= 7;
bitD->bitContainer = MEM_readLEST(bitD->ptr);
return BIT_DStream_unfinished;
return BIT_reloadDStreamFast(bitD);
}
if (bitD->ptr == bitD->start) {
if (bitD->bitsConsumed < sizeof(bitD->bitContainer)*8) return BIT_DStream_endOfBuffer;

78
drivers/filesystems/btrfs/zstd/compiler.h
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -15,13 +15,15 @@
* Compiler specifics
*********************************************************/
/* force inlining */
#if defined (__GNUC__) || defined(__cplusplus) || defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */
#if !defined(ZSTD_NO_INLINE)
#if (defined(__GNUC__) && !defined(__STRICT_ANSI__)) || defined(__cplusplus) || defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */
# define INLINE_KEYWORD inline
#else
# define INLINE_KEYWORD
#endif
#if defined(__GNUC__)
#if defined(__GNUC__) || defined(__ICCARM__)
# define FORCE_INLINE_ATTR __attribute__((always_inline))
#elif defined(_MSC_VER)
# define FORCE_INLINE_ATTR __forceinline
@ -29,9 +31,16 @@
# define FORCE_INLINE_ATTR
#endif
#else
#define INLINE_KEYWORD
#define FORCE_INLINE_ATTR
#endif
/**
* FORCE_INLINE_TEMPLATE is used to define C "templates", which take constant
* parameters. They must be inlined for the compiler to elimininate the constant
* parameters. They must be inlined for the compiler to eliminate the constant
* branches.
*/
#define FORCE_INLINE_TEMPLATE static INLINE_KEYWORD FORCE_INLINE_ATTR
@ -52,11 +61,18 @@
# define HINT_INLINE static INLINE_KEYWORD FORCE_INLINE_ATTR
#endif
/* UNUSED_ATTR tells the compiler it is okay if the function is unused. */
#if defined(__GNUC__)
# define UNUSED_ATTR __attribute__((unused))
#else
# define UNUSED_ATTR
#endif
/* force no inlining */
#ifdef _MSC_VER
# define FORCE_NOINLINE static __declspec(noinline)
#else
# ifdef __GNUC__
# if defined(__GNUC__) || defined(__ICCARM__)
# define FORCE_NOINLINE static __attribute__((__noinline__))
# else
# define FORCE_NOINLINE static
@ -67,7 +83,7 @@
#ifndef __has_attribute
#define __has_attribute(x) 0 /* Compatibility with non-clang compilers. */
#endif
#if defined(__GNUC__)
#if defined(__GNUC__) || defined(__ICCARM__)
# define TARGET_ATTRIBUTE(target) __attribute__((__target__(target)))
#else
# define TARGET_ATTRIBUTE(target)
@ -89,23 +105,24 @@
#endif
/* prefetch
* can be disabled, by declaring NO_PREFETCH macro
* All prefetch invocations use a single default locality 2,
* generating instruction prefetcht1,
* which, according to Intel, means "load data into L2 cache".
* This is a good enough "middle ground" for the time being,
* though in theory, it would be better to specialize locality depending on data being prefetched.
* Tests could not determine any sensible difference based on locality value. */
* can be disabled, by declaring NO_PREFETCH build macro */
#if defined(NO_PREFETCH)
# define PREFETCH(ptr) (void)(ptr) /* disabled */
# define PREFETCH_L1(ptr) (void)(ptr) /* disabled */
# define PREFETCH_L2(ptr) (void)(ptr) /* disabled */
#else
# if defined(_MSC_VER) && (defined(_M_X64) || defined(_M_I86)) /* _mm_prefetch() is not defined outside of x86/x64 */
# include <mmintrin.h> /* https://msdn.microsoft.com/fr-fr/library/84szxsww(v=vs.90).aspx */
# define PREFETCH(ptr) _mm_prefetch((const char*)(ptr), _MM_HINT_T1)
# define PREFETCH_L1(ptr) _mm_prefetch((const char*)(ptr), _MM_HINT_T0)
# define PREFETCH_L2(ptr) _mm_prefetch((const char*)(ptr), _MM_HINT_T1)
# elif defined(__aarch64__)
# define PREFETCH_L1(ptr) __asm__ __volatile__("prfm pldl1keep, %0" ::"Q"(*(ptr)))
# define PREFETCH_L2(ptr) __asm__ __volatile__("prfm pldl2keep, %0" ::"Q"(*(ptr)))
# elif defined(__GNUC__) && ( (__GNUC__ >= 4) || ( (__GNUC__ == 3) && (__GNUC_MINOR__ >= 1) ) )
# define PREFETCH(ptr) __builtin_prefetch((ptr), 0 /* rw==read */, 2 /* locality */)
# define PREFETCH_L1(ptr) __builtin_prefetch((ptr), 0 /* rw==read */, 3 /* locality */)
# define PREFETCH_L2(ptr) __builtin_prefetch((ptr), 0 /* rw==read */, 2 /* locality */)
# else
# define PREFETCH(ptr) (void)(ptr) /* disabled */
# define PREFETCH_L1(ptr) (void)(ptr) /* disabled */
# define PREFETCH_L2(ptr) (void)(ptr) /* disabled */
# endif
#endif /* NO_PREFETCH */
@ -116,10 +133,35 @@
size_t const _size = (size_t)(s); \
size_t _pos; \
for (_pos=0; _pos<_size; _pos+=CACHELINE_SIZE) { \
PREFETCH(_ptr + _pos); \
PREFETCH_L2(_ptr + _pos); \
} \
}
/* vectorization
* older GCC (pre gcc-4.3 picked as the cutoff) uses a different syntax */
#if !defined(__INTEL_COMPILER) && !defined(__clang__) && defined(__GNUC__)
# if (__GNUC__ == 4 && __GNUC_MINOR__ > 3) || (__GNUC__ >= 5)
# define DONT_VECTORIZE __attribute__((optimize("no-tree-vectorize")))
# else
# define DONT_VECTORIZE _Pragma("GCC optimize(\"no-tree-vectorize\")")
# endif
#else
# define DONT_VECTORIZE
#endif
/* Tell the compiler that a branch is likely or unlikely.
* Only use these macros if it causes the compiler to generate better code.
* If you can remove a LIKELY/UNLIKELY annotation without speed changes in gcc
* and clang, please do.
*/
#if defined(__GNUC__)
#define LIKELY(x) (__builtin_expect((x), 1))
#define UNLIKELY(x) (__builtin_expect((x), 0))
#else
#define LIKELY(x) (x)
#define UNLIKELY(x) (x)
#endif
/* disable warnings */
#ifdef _MSC_VER /* Visual Studio */
# include <intrin.h> /* For Visual 2005 */

4
drivers/filesystems/btrfs/zstd/cpu.h
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2018-present, Facebook, Inc.
* Copyright (c) 2018-2020, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -78,7 +78,7 @@ MEM_STATIC ZSTD_cpuid_t ZSTD_cpuid(void) {
__asm__(
"pushl %%ebx\n\t"
"cpuid\n\t"
"movl %%ebx, %%eax\n\r"
"movl %%ebx, %%eax\n\t"
"popl %%ebx"
: "=a"(f7b), "=c"(f7c)
: "a"(7), "c"(0)

69
drivers/filesystems/btrfs/zstd/debug.h
View file

@ -1,35 +1,15 @@
/* ******************************************************************
debug
Part of FSE library
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- Source repository : https://github.com/Cyan4973/FiniteStateEntropy
* debug
* Part of FSE library
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - Source repository : https://github.com/Cyan4973/FiniteStateEntropy
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
@ -57,9 +37,9 @@ extern "C" {
#endif
/* static assert is triggered at compile time, leaving no runtime artefact,
* but can only work with compile-time constants.
* This variant can only be used inside a function. */
/* static assert is triggered at compile time, leaving no runtime artefact.
* static assert only works with compile-time constants.
* Also, this variant can only be used inside a function. */
#define DEBUG_STATIC_ASSERT(c) (void)sizeof(char[(c) ? 1 : -1])
@ -70,9 +50,19 @@ extern "C" {
# define DEBUGLEVEL 0
#endif
/* DEBUGFILE can be defined externally,
* typically through compiler command line.
* note : currently useless.
* Value must be stderr or stdout */
#ifndef DEBUGFILE
# define DEBUGFILE stderr
#endif
/* recommended values for DEBUGLEVEL :
* 0 : no debug, all run-time functions disabled
* 1 : no display, enables assert() only
* 0 : release mode, no debug, all run-time checks disabled
* 1 : enables assert() only, no display
* 2 : reserved, for currently active debug path
* 3 : events once per object lifetime (CCtx, CDict, etc.)
* 4 : events once per frame
@ -81,7 +71,7 @@ extern "C" {
* 7+: events at every position (*very* verbose)
*
* It's generally inconvenient to output traces > 5.
* In which case, it's possible to selectively enable higher verbosity levels
* In which case, it's possible to selectively trigger high verbosity levels
* by modifying g_debug_level.
*/
@ -95,10 +85,11 @@ extern "C" {
#if (DEBUGLEVEL>=2)
# include <stdio.h>
extern int g_debuglevel; /* here, this variable is only declared,
extern int g_debuglevel; /* the variable is only declared,
it actually lives in debug.c,
and is shared by the whole process.
It's typically used to enable very verbose levels
It's not thread-safe.
It's useful when enabling very verbose levels
on selective conditions (such as position in src) */
# define RAWLOG(l, ...) { \

46
drivers/filesystems/btrfs/zstd/entropy_common.c
View file

@ -1,36 +1,16 @@
/*
Common functions of New Generation Entropy library
Copyright (C) 2016, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy
- Public forum : https://groups.google.com/forum/#!forum/lz4c
*************************************************************************** */
/* ******************************************************************
* Common functions of New Generation Entropy library
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy
* - Public forum : https://groups.google.com/forum/#!forum/lz4c
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
/* *************************************
* Dependencies

9
drivers/filesystems/btrfs/zstd/error_private.c
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -14,6 +14,10 @@
const char* ERR_getErrorString(ERR_enum code)
{
#ifdef ZSTD_STRIP_ERROR_STRINGS
(void)code;
return "Error strings stripped";
#else
static const char* const notErrorCode = "Unspecified error code";
switch( code )
{
@ -39,10 +43,13 @@ const char* ERR_getErrorString(ERR_enum code)
case PREFIX(dictionaryCreation_failed): return "Cannot create Dictionary from provided samples";
case PREFIX(dstSize_tooSmall): return "Destination buffer is too small";
case PREFIX(srcSize_wrong): return "Src size is incorrect";
case PREFIX(dstBuffer_null): return "Operation on NULL destination buffer";
/* following error codes are not stable and may be removed or changed in a future version */
case PREFIX(frameIndex_tooLarge): return "Frame index is too large";
case PREFIX(seekableIO): return "An I/O error occurred when reading/seeking";
case PREFIX(dstBuffer_wrong): return "Destination buffer is wrong";
case PREFIX(maxCode):
default: return notErrorCode;
}
#endif
}

8
drivers/filesystems/btrfs/zstd/error_private.h
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -49,7 +49,7 @@ typedef ZSTD_ErrorCode ERR_enum;
/*-****************************************
* Error codes handling
******************************************/
#undef ERROR /* reported already defined on VS 2015 (Rich Geldreich) */
#undef ERROR /* already defined on Visual Studio */
#define ERROR(name) ZSTD_ERROR(name)
#define ZSTD_ERROR(name) ((size_t)-PREFIX(name))
@ -57,6 +57,10 @@ ERR_STATIC unsigned ERR_isError(size_t code) { return (code > ERROR(maxCode)); }
ERR_STATIC ERR_enum ERR_getErrorCode(size_t code) { if (!ERR_isError(code)) return (ERR_enum)0; return (ERR_enum) (0-code); }
/* check and forward error code */
#define CHECK_V_F(e, f) size_t const e = f; if (ERR_isError(e)) return e
#define CHECK_F(f) { CHECK_V_F(_var_err__, f); }
/*-****************************************
* Error Strings

50
drivers/filesystems/btrfs/zstd/fse.h
View file

@ -1,35 +1,15 @@
/* ******************************************************************
FSE : Finite State Entropy codec
Public Prototypes declaration
Copyright (C) 2013-2016, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- Source repository : https://github.com/Cyan4973/FiniteStateEntropy
* FSE : Finite State Entropy codec
* Public Prototypes declaration
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - Source repository : https://github.com/Cyan4973/FiniteStateEntropy
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
#if defined (__cplusplus)
@ -308,7 +288,7 @@ If there is an error, the function will return an error code, which can be teste
*******************************************/
/* FSE buffer bounds */
#define FSE_NCOUNTBOUND 512
#define FSE_BLOCKBOUND(size) (size + (size>>7))
#define FSE_BLOCKBOUND(size) (size + (size>>7) + 4 /* fse states */ + sizeof(size_t) /* bitContainer */)
#define FSE_COMPRESSBOUND(size) (FSE_NCOUNTBOUND + FSE_BLOCKBOUND(size)) /* Macro version, useful for static allocation */
/* It is possible to statically allocate FSE CTable/DTable as a table of FSE_CTable/FSE_DTable using below macros */
@ -358,7 +338,7 @@ size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size
typedef enum {
FSE_repeat_none, /**< Cannot use the previous table */
FSE_repeat_check, /**< Can use the previous table but it must be checked */
FSE_repeat_valid /**< Can use the previous table and it is asumed to be valid */
FSE_repeat_valid /**< Can use the previous table and it is assumed to be valid */
} FSE_repeat;
/* *****************************************
@ -512,7 +492,7 @@ MEM_STATIC void FSE_initCState(FSE_CState_t* statePtr, const FSE_CTable* ct)
const U32 tableLog = MEM_read16(ptr);
statePtr->value = (ptrdiff_t)1<<tableLog;
statePtr->stateTable = u16ptr+2;
statePtr->symbolTT = ((const U32*)ct + 1 + (tableLog ? (1<<(tableLog-1)) : 1));
statePtr->symbolTT = ct + 1 + (tableLog ? (1<<(tableLog-1)) : 1);
statePtr->stateLog = tableLog;
}
@ -531,7 +511,7 @@ MEM_STATIC void FSE_initCState2(FSE_CState_t* statePtr, const FSE_CTable* ct, U3
}
}
MEM_STATIC void FSE_encodeSymbol(BIT_CStream_t* bitC, FSE_CState_t* statePtr, U32 symbol)
MEM_STATIC void FSE_encodeSymbol(BIT_CStream_t* bitC, FSE_CState_t* statePtr, unsigned symbol)
{
FSE_symbolCompressionTransform const symbolTT = ((const FSE_symbolCompressionTransform*)(statePtr->symbolTT))[symbol];
const U16* const stateTable = (const U16*)(statePtr->stateTable);

90
drivers/filesystems/btrfs/zstd/fse_compress.c
View file

@ -1,35 +1,15 @@
/* ******************************************************************
FSE : Finite State Entropy encoder
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
- Public forum : https://groups.google.com/forum/#!forum/lz4c
* FSE : Finite State Entropy encoder
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
* - Public forum : https://groups.google.com/forum/#!forum/lz4c
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
/* **************************************************************
@ -118,7 +98,7 @@ size_t FSE_buildCTable_wksp(FSE_CTable* ct,
/* symbol start positions */
{ U32 u;
cumul[0] = 0;
for (u=1; u<=maxSymbolValue+1; u++) {
for (u=1; u <= maxSymbolValue+1; u++) {
if (normalizedCounter[u-1]==-1) { /* Low proba symbol */
cumul[u] = cumul[u-1] + 1;
tableSymbol[highThreshold--] = (FSE_FUNCTION_TYPE)(u-1);
@ -132,9 +112,9 @@ size_t FSE_buildCTable_wksp(FSE_CTable* ct,
{ U32 position = 0;
U32 symbol;
for (symbol=0; symbol<=maxSymbolValue; symbol++) {
int nbOccurences;
int nbOccurrences;
int const freq = normalizedCounter[symbol];
for (nbOccurences=0; nbOccurences<freq; nbOccurences++) {
for (nbOccurrences=0; nbOccurrences<freq; nbOccurrences++) {
tableSymbol[position] = (FSE_FUNCTION_TYPE)symbol;
position = (position + step) & tableMask;
while (position > highThreshold)
@ -194,17 +174,8 @@ size_t FSE_buildCTable_wksp(FSE_CTable* ct,
size_t FSE_buildCTable(FSE_CTable* ct, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog)
{
FSE_FUNCTION_TYPE* tableSymbol = ExAllocatePoolWithTag(NonPagedPool, sizeof(FSE_FUNCTION_TYPE) * FSE_MAX_TABLESIZE, FSEC_ALLOC_TAG);
size_t ret;
if (!tableSymbol)
return 0;
ret = FSE_buildCTable_wksp(ct, normalizedCounter, maxSymbolValue, tableLog, tableSymbol, sizeof(FSE_FUNCTION_TYPE) * FSE_MAX_TABLESIZE);
ExFreePool(tableSymbol);
return ret;
FSE_FUNCTION_TYPE tableSymbol[FSE_MAX_TABLESIZE]; /* memset() is not necessary, even if static analyzer complain about it */
return FSE_buildCTable_wksp(ct, normalizedCounter, maxSymbolValue, tableLog, tableSymbol, sizeof(tableSymbol));
}
@ -657,9 +628,6 @@ size_t FSE_compress_usingCTable (void* dst, size_t dstSize,
size_t FSE_compressBound(size_t size) { return FSE_COMPRESSBOUND(size); }
#define CHECK_V_F(e, f) size_t const e = f; if (ERR_isError(e)) return e
#define CHECK_F(f) { CHECK_V_F(_var_err__, f); }
/* FSE_compress_wksp() :
* Same as FSE_compress2(), but using an externally allocated scratch buffer (`workSpace`).
* `wkspSize` size must be `(1<<tableLog)`.
@ -670,7 +638,7 @@ size_t FSE_compress_wksp (void* dst, size_t dstSize, const void* src, size_t src
BYTE* op = ostart;
BYTE* const oend = ostart + dstSize;
U32 count[FSE_MAX_SYMBOL_VALUE+1];
unsigned count[FSE_MAX_SYMBOL_VALUE+1];
S16 norm[FSE_MAX_SYMBOL_VALUE+1];
FSE_CTable* CTable = (FSE_CTable*)workSpace;
size_t const CTableSize = FSE_CTABLE_SIZE_U32(tableLog, maxSymbolValue);
@ -684,7 +652,7 @@ size_t FSE_compress_wksp (void* dst, size_t dstSize, const void* src, size_t src
if (!tableLog) tableLog = FSE_DEFAULT_TABLELOG;
/* Scan input and build symbol stats */
{ CHECK_V_F(maxCount, HIST_count_wksp(count, &maxSymbolValue, src, srcSize, (unsigned*)scratchBuffer) );
{ CHECK_V_F(maxCount, HIST_count_wksp(count, &maxSymbolValue, src, srcSize, scratchBuffer, scratchBufferSize) );
if (maxCount == srcSize) return 1; /* only a single symbol in src : rle */
if (maxCount == 1) return 0; /* each symbol present maximum once => not compressible */
if (maxCount < (srcSize >> 7)) return 0; /* Heuristic : not compressible enough */
@ -718,24 +686,10 @@ typedef struct {
size_t FSE_compress2 (void* dst, size_t dstCapacity, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog)
{
fseWkspMax_t* scratchBuffer;
size_t ret;
DEBUG_STATIC_ASSERT(sizeof(fseWkspMax_t) >= FSE_WKSP_SIZE_U32(FSE_MAX_TABLELOG, FSE_MAX_SYMBOL_VALUE)); /* compilation failures here means scratchBuffer is not large enough */
if (tableLog > FSE_MAX_TABLELOG)
return ERROR(tableLog_tooLarge);
scratchBuffer = ExAllocatePoolWithTag(NonPagedPool, sizeof(fseWkspMax_t), FSEC_ALLOC_TAG);
if (!scratchBuffer)
return 0;
ret = FSE_compress_wksp(dst, dstCapacity, src, srcSize, maxSymbolValue, tableLog, scratchBuffer, sizeof(fseWkspMax_t));
ExFreePool(scratchBuffer);
return ret;
fseWkspMax_t scratchBuffer;
DEBUG_STATIC_ASSERT(sizeof(scratchBuffer) >= FSE_WKSP_SIZE_U32(FSE_MAX_TABLELOG, FSE_MAX_SYMBOL_VALUE)); /* compilation failures here means scratchBuffer is not large enough */
if (tableLog > FSE_MAX_TABLELOG) return ERROR(tableLog_tooLarge);
return FSE_compress_wksp(dst, dstCapacity, src, srcSize, maxSymbolValue, tableLog, &scratchBuffer, sizeof(scratchBuffer));
}
size_t FSE_compress (void* dst, size_t dstCapacity, const void* src, size_t srcSize)

48
drivers/filesystems/btrfs/zstd/fse_decompress.c
View file

@ -1,35 +1,15 @@
/* ******************************************************************
FSE : Finite State Entropy decoder
Copyright (C) 2013-2015, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
- Public forum : https://groups.google.com/forum/#!forum/lz4c
* FSE : Finite State Entropy decoder
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
* - Public forum : https://groups.google.com/forum/#!forum/lz4c
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
@ -53,9 +33,6 @@
#define FSE_isError ERR_isError
#define FSE_STATIC_ASSERT(c) DEBUG_STATIC_ASSERT(c) /* use only *after* variable declarations */
/* check and forward error code */
#define CHECK_F(f) { size_t const e = f; if (FSE_isError(e)) return e; }
/* **************************************************************
* Templates
@ -81,7 +58,6 @@
#define FSED_ALLOC_TAG 0x64455346 // "FSEd"
/* Function templates */
FSE_DTable* FSE_createDTable (unsigned tableLog)
{
@ -289,7 +265,7 @@ size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size
/* normal FSE decoding mode */
size_t const NCountLength = FSE_readNCount (counting, &maxSymbolValue, &tableLog, istart, cSrcSize);
if (FSE_isError(NCountLength)) return NCountLength;
//if (NCountLength >= cSrcSize) return ERROR(srcSize_wrong); /* too small input size; supposed to be already checked in NCountLength, only remaining case : NCountLength==cSrcSize */
/* if (NCountLength >= cSrcSize) return ERROR(srcSize_wrong); */ /* too small input size; supposed to be already checked in NCountLength, only remaining case : NCountLength==cSrcSize */
if (tableLog > maxLog) return ERROR(tableLog_tooLarge);
ip += NCountLength;
cSrcSize -= NCountLength;

88
drivers/filesystems/btrfs/zstd/hist.c
View file

@ -1,36 +1,16 @@
/* ******************************************************************
hist : Histogram functions
part of Finite State Entropy project
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
- Public forum : https://groups.google.com/forum/#!forum/lz4c
* hist : Histogram functions
* part of Finite State Entropy project
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
* - Public forum : https://groups.google.com/forum/#!forum/lz4c
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
/* --- dependencies --- */
@ -38,10 +18,7 @@
#include "debug.h" /* assert, DEBUGLOG */
#include "error_private.h" /* ERROR */
#include "hist.h"
#include <ntifs.h>
#include <ntddk.h>
#define HIST_ALLOC_TAG 0x54534948 // "HIST"
/* --- Error management --- */
unsigned HIST_isError(size_t code) { return ERR_isError(code); }
@ -76,6 +53,7 @@ unsigned HIST_count_simple(unsigned* count, unsigned* maxSymbolValuePtr,
return largestCount;
}
typedef enum { trustInput, checkMaxSymbolValue } HIST_checkInput_e;
/* HIST_count_parallel_wksp() :
* store histogram into 4 intermediate tables, recombined at the end.
@ -88,8 +66,8 @@ unsigned HIST_count_simple(unsigned* count, unsigned* maxSymbolValuePtr,
static size_t HIST_count_parallel_wksp(
unsigned* count, unsigned* maxSymbolValuePtr,
const void* source, size_t sourceSize,
unsigned checkMax,
unsigned* const workSpace)
HIST_checkInput_e check,
U32* const workSpace)
{
const BYTE* ip = (const BYTE*)source;
const BYTE* const iend = ip+sourceSize;
@ -140,7 +118,7 @@ static size_t HIST_count_parallel_wksp(
/* finish last symbols */
while (ip<iend) Counting1[*ip++]++;
if (checkMax) { /* verify stats will fit into destination table */
if (check) { /* verify stats will fit into destination table */
U32 s; for (s=255; s>maxSymbolValue; s--) {
Counting1[s] += Counting2[s] + Counting3[s] + Counting4[s];
if (Counting1[s]) return ERROR(maxSymbolValue_tooSmall);
@ -160,48 +138,46 @@ static size_t HIST_count_parallel_wksp(
/* HIST_countFast_wksp() :
* Same as HIST_countFast(), but using an externally provided scratch buffer.
* `workSpace` size must be table of >= HIST_WKSP_SIZE_U32 unsigned */
* `workSpace` is a writable buffer which must be 4-bytes aligned,
* `workSpaceSize` must be >= HIST_WKSP_SIZE
*/
size_t HIST_countFast_wksp(unsigned* count, unsigned* maxSymbolValuePtr,
const void* source, size_t sourceSize,
unsigned* workSpace)
void* workSpace, size_t workSpaceSize)
{
if (sourceSize < 1500) /* heuristic threshold */
return HIST_count_simple(count, maxSymbolValuePtr, source, sourceSize);
return HIST_count_parallel_wksp(count, maxSymbolValuePtr, source, sourceSize, 0, workSpace);
if ((size_t)workSpace & 3) return ERROR(GENERIC); /* must be aligned on 4-bytes boundaries */
if (workSpaceSize < HIST_WKSP_SIZE) return ERROR(workSpace_tooSmall);
return HIST_count_parallel_wksp(count, maxSymbolValuePtr, source, sourceSize, trustInput, (U32*)workSpace);
}
/* fast variant (unsafe : won't check if src contains values beyond count[] limit) */
size_t HIST_countFast(unsigned* count, unsigned* maxSymbolValuePtr,
const void* source, size_t sourceSize)
{
unsigned* tmpCounters = ExAllocatePoolWithTag(NonPagedPool, sizeof(unsigned) * HIST_WKSP_SIZE_U32, HIST_ALLOC_TAG);
size_t ret;
if (!tmpCounters)
return 0;
ret = HIST_countFast_wksp(count, maxSymbolValuePtr, source, sourceSize, tmpCounters);
ExFreePool(tmpCounters);
return ret;
unsigned tmpCounters[HIST_WKSP_SIZE_U32];
return HIST_countFast_wksp(count, maxSymbolValuePtr, source, sourceSize, tmpCounters, sizeof(tmpCounters));
}
/* HIST_count_wksp() :
* Same as HIST_count(), but using an externally provided scratch buffer.
* `workSpace` size must be table of >= HIST_WKSP_SIZE_U32 unsigned */
size_t HIST_count_wksp(unsigned* count, unsigned* maxSymbolValuePtr,
const void* source, size_t sourceSize, unsigned* workSpace)
const void* source, size_t sourceSize,
void* workSpace, size_t workSpaceSize)
{
if ((size_t)workSpace & 3) return ERROR(GENERIC); /* must be aligned on 4-bytes boundaries */
if (workSpaceSize < HIST_WKSP_SIZE) return ERROR(workSpace_tooSmall);
if (*maxSymbolValuePtr < 255)
return HIST_count_parallel_wksp(count, maxSymbolValuePtr, source, sourceSize, 1, workSpace);
return HIST_count_parallel_wksp(count, maxSymbolValuePtr, source, sourceSize, checkMaxSymbolValue, (U32*)workSpace);
*maxSymbolValuePtr = 255;
return HIST_countFast_wksp(count, maxSymbolValuePtr, source, sourceSize, workSpace);
return HIST_countFast_wksp(count, maxSymbolValuePtr, source, sourceSize, workSpace, workSpaceSize);
}
size_t HIST_count(unsigned* count, unsigned* maxSymbolValuePtr,
const void* src, size_t srcSize)
{
unsigned tmpCounters[HIST_WKSP_SIZE_U32];
return HIST_count_wksp(count, maxSymbolValuePtr, src, srcSize, tmpCounters);
return HIST_count_wksp(count, maxSymbolValuePtr, src, srcSize, tmpCounters, sizeof(tmpCounters));
}

55
drivers/filesystems/btrfs/zstd/hist.h
View file

@ -1,36 +1,16 @@
/* ******************************************************************
hist : Histogram functions
part of Finite State Entropy project
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
- Public forum : https://groups.google.com/forum/#!forum/lz4c
* hist : Histogram functions
* part of Finite State Entropy project
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - FSE source repository : https://github.com/Cyan4973/FiniteStateEntropy
* - Public forum : https://groups.google.com/forum/#!forum/lz4c
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
/* --- dependencies --- */
@ -56,14 +36,16 @@ unsigned HIST_isError(size_t code); /**< tells if a return value is an error co
/* --- advanced histogram functions --- */
#define HIST_WKSP_SIZE_U32 1024
#define HIST_WKSP_SIZE (HIST_WKSP_SIZE_U32 * sizeof(unsigned))
/** HIST_count_wksp() :
* Same as HIST_count(), but using an externally provided scratch buffer.
* Benefit is this function will use very little stack space.
* `workSpace` must be a table of unsigned of size >= HIST_WKSP_SIZE_U32
* `workSpace` is a writable buffer which must be 4-bytes aligned,
* `workSpaceSize` must be >= HIST_WKSP_SIZE
*/
size_t HIST_count_wksp(unsigned* count, unsigned* maxSymbolValuePtr,
const void* src, size_t srcSize,
unsigned* workSpace);
void* workSpace, size_t workSpaceSize);
/** HIST_countFast() :
* same as HIST_count(), but blindly trusts that all byte values within src are <= *maxSymbolValuePtr.
@ -74,11 +56,12 @@ size_t HIST_countFast(unsigned* count, unsigned* maxSymbolValuePtr,
/** HIST_countFast_wksp() :
* Same as HIST_countFast(), but using an externally provided scratch buffer.
* `workSpace` must be a table of unsigned of size >= HIST_WKSP_SIZE_U32
* `workSpace` is a writable buffer which must be 4-bytes aligned,
* `workSpaceSize` must be >= HIST_WKSP_SIZE
*/
size_t HIST_countFast_wksp(unsigned* count, unsigned* maxSymbolValuePtr,
const void* src, size_t srcSize,
unsigned* workSpace);
void* workSpace, size_t workSpaceSize);
/*! HIST_count_simple() :
* Same as HIST_countFast(), this function is unsafe,

78
drivers/filesystems/btrfs/zstd/huf.h
View file

@ -1,35 +1,15 @@
/* ******************************************************************
huff0 huffman codec,
part of Finite State Entropy library
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- Source repository : https://github.com/Cyan4973/FiniteStateEntropy
* huff0 huffman codec,
* part of Finite State Entropy library
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - Source repository : https://github.com/Cyan4973/FiniteStateEntropy
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
#if defined (__cplusplus)
@ -110,7 +90,7 @@ HUF_PUBLIC_API size_t HUF_compress2 (void* dst, size_t dstCapacity,
/** HUF_compress4X_wksp() :
* Same as HUF_compress2(), but uses externally allocated `workSpace`.
* `workspace` must have minimum alignment of 4, and be at least as large as HUF_WORKSPACE_SIZE */
#define HUF_WORKSPACE_SIZE (6 << 10)
#define HUF_WORKSPACE_SIZE ((6 << 10) + 256)
#define HUF_WORKSPACE_SIZE_U32 (HUF_WORKSPACE_SIZE / sizeof(U32))
HUF_PUBLIC_API size_t HUF_compress4X_wksp (void* dst, size_t dstCapacity,
const void* src, size_t srcSize,
@ -163,21 +143,29 @@ HUF_PUBLIC_API size_t HUF_compress4X_wksp (void* dst, size_t dstCapacity,
/* static allocation of HUF's DTable */
typedef U32 HUF_DTable;
#define HUF_DTABLE_SIZE(maxTableLog) (1 + (1<<(maxTableLog)))
#define HUF_CREATE_STATIC_DTABLEX1(DTable, maxTableLog) \
HUF_DTable DTable[HUF_DTABLE_SIZE((maxTableLog)-1)] = { ((U32)((maxTableLog)-1) * 0x01000001) }
#define HUF_CREATE_STATIC_DTABLEX2(DTable, maxTableLog) \
HUF_DTable DTable[HUF_DTABLE_SIZE(maxTableLog)] = { ((U32)(maxTableLog) * 0x01000001) }
/* ****************************************
* Advanced decompression functions
******************************************/
size_t HUF_decompress4X1 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< single-symbol decoder */
#ifndef HUF_FORCE_DECOMPRESS_X1
size_t HUF_decompress4X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< double-symbols decoder */
#endif
size_t HUF_decompress4X_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< decodes RLE and uncompressed */
size_t HUF_decompress4X_hufOnly(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< considers RLE and uncompressed as errors */
size_t HUF_decompress4X_hufOnly_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< considers RLE and uncompressed as errors */
size_t HUF_decompress4X1_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< single-symbol decoder */
size_t HUF_decompress4X1_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< single-symbol decoder */
#ifndef HUF_FORCE_DECOMPRESS_X1
size_t HUF_decompress4X2_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< double-symbols decoder */
size_t HUF_decompress4X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< double-symbols decoder */
#endif
/* ****************************************
@ -200,6 +188,8 @@ typedef struct HUF_CElt_s HUF_CElt; /* incomplete type */
size_t HUF_buildCTable (HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue, unsigned maxNbBits); /* @return : maxNbBits; CTable and count can overlap. In which case, CTable will overwrite count content */
size_t HUF_writeCTable (void* dst, size_t maxDstSize, const HUF_CElt* CTable, unsigned maxSymbolValue, unsigned huffLog);
size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable);
size_t HUF_estimateCompressedSize(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue);
int HUF_validateCTable(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue);
typedef enum {
HUF_repeat_none, /**< Cannot use the previous table */
@ -224,7 +214,7 @@ size_t HUF_compress4X_repeat(void* dst, size_t dstSize,
#define HUF_CTABLE_WORKSPACE_SIZE_U32 (2*HUF_SYMBOLVALUE_MAX +1 +1)
#define HUF_CTABLE_WORKSPACE_SIZE (HUF_CTABLE_WORKSPACE_SIZE_U32 * sizeof(unsigned))
size_t HUF_buildCTable_wksp (HUF_CElt* tree,
const U32* count, U32 maxSymbolValue, U32 maxNbBits,
const unsigned* count, U32 maxSymbolValue, U32 maxNbBits,
void* workSpace, size_t wkspSize);
/*! HUF_readStats() :
@ -238,7 +228,7 @@ size_t HUF_readStats(BYTE* huffWeight, size_t hwSize,
/** HUF_readCTable() :
* Loading a CTable saved with HUF_writeCTable() */
size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize);
size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize, unsigned *hasZeroWeights);
/** HUF_getNbBits() :
* Read nbBits from CTable symbolTable, for symbol `symbolValue` presumed <= HUF_SYMBOLVALUE_MAX
@ -273,14 +263,22 @@ U32 HUF_selectDecoder (size_t dstSize, size_t cSrcSize);
#define HUF_DECOMPRESS_WORKSPACE_SIZE (2 << 10)
#define HUF_DECOMPRESS_WORKSPACE_SIZE_U32 (HUF_DECOMPRESS_WORKSPACE_SIZE / sizeof(U32))
#ifndef HUF_FORCE_DECOMPRESS_X2
size_t HUF_readDTableX1 (HUF_DTable* DTable, const void* src, size_t srcSize);
size_t HUF_readDTableX1_wksp (HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize);
#endif
#ifndef HUF_FORCE_DECOMPRESS_X1
size_t HUF_readDTableX2 (HUF_DTable* DTable, const void* src, size_t srcSize);
size_t HUF_readDTableX2_wksp (HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize);
#endif
size_t HUF_decompress4X_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
#ifndef HUF_FORCE_DECOMPRESS_X2
size_t HUF_decompress4X1_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
#endif
#ifndef HUF_FORCE_DECOMPRESS_X1
size_t HUF_decompress4X2_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
#endif
/* ====================== */
@ -302,24 +300,36 @@ size_t HUF_compress1X_repeat(void* dst, size_t dstSize,
HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2);
size_t HUF_decompress1X1 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /* single-symbol decoder */
#ifndef HUF_FORCE_DECOMPRESS_X1
size_t HUF_decompress1X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /* double-symbol decoder */
#endif
size_t HUF_decompress1X_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);
size_t HUF_decompress1X_DCtx_wksp (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize);
#ifndef HUF_FORCE_DECOMPRESS_X2
size_t HUF_decompress1X1_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< single-symbol decoder */
size_t HUF_decompress1X1_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< single-symbol decoder */
#endif
#ifndef HUF_FORCE_DECOMPRESS_X1
size_t HUF_decompress1X2_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< double-symbols decoder */
size_t HUF_decompress1X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< double-symbols decoder */
#endif
size_t HUF_decompress1X_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable); /**< automatic selection of sing or double symbol decoder, based on DTable */
#ifndef HUF_FORCE_DECOMPRESS_X2
size_t HUF_decompress1X1_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
#endif
#ifndef HUF_FORCE_DECOMPRESS_X1
size_t HUF_decompress1X2_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable);
#endif
/* BMI2 variants.
* If the CPU has BMI2 support, pass bmi2=1, otherwise pass bmi2=0.
*/
size_t HUF_decompress1X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2);
#ifndef HUF_FORCE_DECOMPRESS_X2
size_t HUF_decompress1X1_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2);
#endif
size_t HUF_decompress4X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2);
size_t HUF_decompress4X_hufOnly_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2);

259
drivers/filesystems/btrfs/zstd/huf_compress.c
View file

@ -1,35 +1,15 @@
/* ******************************************************************
Huffman encoder, part of New Generation Entropy library
Copyright (C) 2013-2016, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy
- Public forum : https://groups.google.com/forum/#!forum/lz4c
* Huffman encoder, part of New Generation Entropy library
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy
* - Public forum : https://groups.google.com/forum/#!forum/lz4c
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
/* **************************************************************
@ -53,10 +33,6 @@
#define HUF_STATIC_LINKING_ONLY
#include "huf.h"
#include "error_private.h"
#include <ntifs.h>
#include <ntddk.h>
#define HUFC_ALLOC_TAG 0x63465548 // "HUFc"
/* **************************************************************
@ -64,8 +40,6 @@
****************************************************************/
#define HUF_isError ERR_isError
#define HUF_STATIC_ASSERT(c) DEBUG_STATIC_ASSERT(c) /* use only *after* variable declarations */
#define CHECK_V_F(e, f) size_t const e = f; if (ERR_isError(e)) return e
#define CHECK_F(f) { CHECK_V_F(_var_err__, f); }
/* **************************************************************
@ -92,13 +66,13 @@ static size_t HUF_compressWeights (void* dst, size_t dstSize, const void* weight
BYTE* op = ostart;
BYTE* const oend = ostart + dstSize;
U32 maxSymbolValue = HUF_TABLELOG_MAX;
unsigned maxSymbolValue = HUF_TABLELOG_MAX;
U32 tableLog = MAX_FSE_TABLELOG_FOR_HUFF_HEADER;
FSE_CTable CTable[FSE_CTABLE_SIZE_U32(MAX_FSE_TABLELOG_FOR_HUFF_HEADER, HUF_TABLELOG_MAX)];
BYTE scratchBuffer[1<<MAX_FSE_TABLELOG_FOR_HUFF_HEADER];
U32 count[HUF_TABLELOG_MAX+1];
unsigned count[HUF_TABLELOG_MAX+1];
S16 norm[HUF_TABLELOG_MAX+1];
/* init conditions */
@ -114,18 +88,18 @@ static size_t HUF_compressWeights (void* dst, size_t dstSize, const void* weight
CHECK_F( FSE_normalizeCount(norm, tableLog, count, wtSize, maxSymbolValue) );
/* Write table description header */
{ CHECK_V_F(hSize, FSE_writeNCount(op, oend-op, norm, maxSymbolValue, tableLog) );
{ CHECK_V_F(hSize, FSE_writeNCount(op, (size_t)(oend-op), norm, maxSymbolValue, tableLog) );
op += hSize;
}
/* Compress */
CHECK_F( FSE_buildCTable_wksp(CTable, norm, maxSymbolValue, tableLog, scratchBuffer, sizeof(scratchBuffer)) );
{ CHECK_V_F(cSize, FSE_compress_usingCTable(op, oend - op, weightTable, wtSize, CTable) );
{ CHECK_V_F(cSize, FSE_compress_usingCTable(op, (size_t)(oend - op), weightTable, wtSize, CTable) );
if (cSize == 0) return 0; /* not enough space for compressed data */
op += cSize;
}
return op-ostart;
return (size_t)(op-ostart);
}
@ -138,7 +112,7 @@ struct HUF_CElt_s {
`CTable` : Huffman tree to save, using huf representation.
@return : size of saved CTable */
size_t HUF_writeCTable (void* dst, size_t maxDstSize,
const HUF_CElt* CTable, U32 maxSymbolValue, U32 huffLog)
const HUF_CElt* CTable, unsigned maxSymbolValue, unsigned huffLog)
{
BYTE bitsToWeight[HUF_TABLELOG_MAX + 1]; /* precomputed conversion table */
BYTE huffWeight[HUF_SYMBOLVALUE_MAX];
@ -173,7 +147,7 @@ size_t HUF_writeCTable (void* dst, size_t maxDstSize,
}
size_t HUF_readCTable (HUF_CElt* CTable, U32* maxSymbolValuePtr, const void* src, size_t srcSize)
size_t HUF_readCTable (HUF_CElt* CTable, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize, unsigned* hasZeroWeights)
{
BYTE huffWeight[HUF_SYMBOLVALUE_MAX + 1]; /* init not required, even though some static analyzer may complain */
U32 rankVal[HUF_TABLELOG_ABSOLUTEMAX + 1]; /* large enough for values from 0 to 16 */
@ -196,9 +170,11 @@ size_t HUF_readCTable (HUF_CElt* CTable, U32* maxSymbolValuePtr, const void* src
} }
/* fill nbBits */
*hasZeroWeights = 0;
{ U32 n; for (n=0; n<nbSymbols; n++) {
const U32 w = huffWeight[n];
CTable[n].nbBits = (BYTE)(tableLog + 1 - w);
*hasZeroWeights |= (w == 0);
CTable[n].nbBits = (BYTE)(tableLog + 1 - w) & -(w != 0);
} }
/* fill val */
@ -244,7 +220,7 @@ static U32 HUF_setMaxHeight(nodeElt* huffNode, U32 lastNonNull, U32 maxNbBits)
/* there are several too large elements (at least >= 2) */
{ int totalCost = 0;
const U32 baseCost = 1 << (largestBits - maxNbBits);
U32 n = lastNonNull;
int n = (int)lastNonNull;
while (huffNode[n].nbBits > maxNbBits) {
totalCost += baseCost - (1 << (largestBits - huffNode[n].nbBits));
@ -259,22 +235,22 @@ static U32 HUF_setMaxHeight(nodeElt* huffNode, U32 lastNonNull, U32 maxNbBits)
/* repay normalized cost */
{ U32 const noSymbol = 0xF0F0F0F0;
U32 rankLast[HUF_TABLELOG_MAX+2];
int pos;
/* Get pos of last (smallest) symbol per rank */
memset(rankLast, 0xF0, sizeof(rankLast));
{ U32 currentNbBits = maxNbBits;
int pos;
for (pos=n ; pos >= 0; pos--) {
if (huffNode[pos].nbBits >= currentNbBits) continue;
currentNbBits = huffNode[pos].nbBits; /* < maxNbBits */
rankLast[maxNbBits-currentNbBits] = pos;
rankLast[maxNbBits-currentNbBits] = (U32)pos;
} }
while (totalCost > 0) {
U32 nBitsToDecrease = BIT_highbit32(totalCost) + 1;
U32 nBitsToDecrease = BIT_highbit32((U32)totalCost) + 1;
for ( ; nBitsToDecrease > 1; nBitsToDecrease--) {
U32 highPos = rankLast[nBitsToDecrease];
U32 lowPos = rankLast[nBitsToDecrease-1];
U32 const highPos = rankLast[nBitsToDecrease];
U32 const lowPos = rankLast[nBitsToDecrease-1];
if (highPos == noSymbol) continue;
if (lowPos == noSymbol) break;
{ U32 const highTotal = huffNode[highPos].count;
@ -301,7 +277,8 @@ static U32 HUF_setMaxHeight(nodeElt* huffNode, U32 lastNonNull, U32 maxNbBits)
if (rankLast[1] == noSymbol) { /* special case : no rank 1 symbol (using maxNbBits-1); let's create one from largest rank 0 (using maxNbBits) */
while (huffNode[n].nbBits == maxNbBits) n--;
huffNode[n+1].nbBits--;
rankLast[1] = n+1;
assert(n >= 0);
rankLast[1] = (U32)(n+1);
totalCost++;
continue;
}
@ -313,29 +290,36 @@ static U32 HUF_setMaxHeight(nodeElt* huffNode, U32 lastNonNull, U32 maxNbBits)
return maxNbBits;
}
typedef struct {
U32 base;
U32 current;
} rankPos;
static void HUF_sort(nodeElt* huffNode, const U32* count, U32 maxSymbolValue)
typedef nodeElt huffNodeTable[HUF_CTABLE_WORKSPACE_SIZE_U32];
#define RANK_POSITION_TABLE_SIZE 32
typedef struct {
huffNodeTable huffNodeTbl;
rankPos rankPosition[RANK_POSITION_TABLE_SIZE];
} HUF_buildCTable_wksp_tables;
static void HUF_sort(nodeElt* huffNode, const unsigned* count, U32 maxSymbolValue, rankPos* rankPosition)
{
rankPos rank[32];
U32 n;
memset(rank, 0, sizeof(rank));
memset(rankPosition, 0, sizeof(*rankPosition) * RANK_POSITION_TABLE_SIZE);
for (n=0; n<=maxSymbolValue; n++) {
U32 r = BIT_highbit32(count[n] + 1);
rank[r].base ++;
rankPosition[r].base ++;
}
for (n=30; n>0; n--) rank[n-1].base += rank[n].base;
for (n=0; n<32; n++) rank[n].current = rank[n].base;
for (n=30; n>0; n--) rankPosition[n-1].base += rankPosition[n].base;
for (n=0; n<32; n++) rankPosition[n].current = rankPosition[n].base;
for (n=0; n<=maxSymbolValue; n++) {
U32 const c = count[n];
U32 const r = BIT_highbit32(c+1) + 1;
U32 pos = rank[r].current++;
while ((pos > rank[r].base) && (c > huffNode[pos-1].count)) {
U32 pos = rankPosition[r].current++;
while ((pos > rankPosition[r].base) && (c > huffNode[pos-1].count)) {
huffNode[pos] = huffNode[pos-1];
pos--;
}
@ -347,45 +331,48 @@ static void HUF_sort(nodeElt* huffNode, const U32* count, U32 maxSymbolValue)
/** HUF_buildCTable_wksp() :
* Same as HUF_buildCTable(), but using externally allocated scratch buffer.
* `workSpace` must be aligned on 4-bytes boundaries, and be at least as large as a table of HUF_CTABLE_WORKSPACE_SIZE_U32 unsigned.
* `workSpace` must be aligned on 4-bytes boundaries, and be at least as large as sizeof(HUF_buildCTable_wksp_tables).
*/
#define STARTNODE (HUF_SYMBOLVALUE_MAX+1)
typedef nodeElt huffNodeTable[HUF_CTABLE_WORKSPACE_SIZE_U32];
size_t HUF_buildCTable_wksp (HUF_CElt* tree, const U32* count, U32 maxSymbolValue, U32 maxNbBits, void* workSpace, size_t wkspSize)
size_t HUF_buildCTable_wksp (HUF_CElt* tree, const unsigned* count, U32 maxSymbolValue, U32 maxNbBits, void* workSpace, size_t wkspSize)
{
nodeElt* const huffNode0 = (nodeElt*)workSpace;
HUF_buildCTable_wksp_tables* const wksp_tables = (HUF_buildCTable_wksp_tables*)workSpace;
nodeElt* const huffNode0 = wksp_tables->huffNodeTbl;
nodeElt* const huffNode = huffNode0+1;
U32 n, nonNullRank;
int nonNullRank;
int lowS, lowN;
U16 nodeNb = STARTNODE;
U32 nodeRoot;
int nodeNb = STARTNODE;
int n, nodeRoot;
/* safety checks */
if (((size_t)workSpace & 3) != 0) return ERROR(GENERIC); /* must be aligned on 4-bytes boundaries */
if (wkspSize < sizeof(huffNodeTable)) return ERROR(workSpace_tooSmall);
if (wkspSize < sizeof(HUF_buildCTable_wksp_tables))
return ERROR(workSpace_tooSmall);
if (maxNbBits == 0) maxNbBits = HUF_TABLELOG_DEFAULT;
if (maxSymbolValue > HUF_SYMBOLVALUE_MAX) return ERROR(maxSymbolValue_tooLarge);
if (maxSymbolValue > HUF_SYMBOLVALUE_MAX)
return ERROR(maxSymbolValue_tooLarge);
memset(huffNode0, 0, sizeof(huffNodeTable));
/* sort, decreasing order */
HUF_sort(huffNode, count, maxSymbolValue);
HUF_sort(huffNode, count, maxSymbolValue, wksp_tables->rankPosition);
/* init for parents */
nonNullRank = maxSymbolValue;
nonNullRank = (int)maxSymbolValue;
while(huffNode[nonNullRank].count == 0) nonNullRank--;
lowS = nonNullRank; nodeRoot = nodeNb + lowS - 1; lowN = nodeNb;
huffNode[nodeNb].count = huffNode[lowS].count + huffNode[lowS-1].count;
huffNode[lowS].parent = huffNode[lowS-1].parent = nodeNb;
huffNode[lowS].parent = huffNode[lowS-1].parent = (U16)nodeNb;
nodeNb++; lowS-=2;
for (n=nodeNb; n<=nodeRoot; n++) huffNode[n].count = (U32)(1U<<30);
huffNode0[0].count = (U32)(1U<<31); /* fake entry, strong barrier */
/* create parents */
while (nodeNb <= nodeRoot) {
U32 n1 = (huffNode[lowS].count < huffNode[lowN].count) ? lowS-- : lowN++;
U32 n2 = (huffNode[lowS].count < huffNode[lowN].count) ? lowS-- : lowN++;
int const n1 = (huffNode[lowS].count < huffNode[lowN].count) ? lowS-- : lowN++;
int const n2 = (huffNode[lowS].count < huffNode[lowN].count) ? lowS-- : lowN++;
huffNode[nodeNb].count = huffNode[n1].count + huffNode[n2].count;
huffNode[n1].parent = huffNode[n2].parent = nodeNb;
huffNode[n1].parent = huffNode[n2].parent = (U16)nodeNb;
nodeNb++;
}
@ -397,24 +384,25 @@ size_t HUF_buildCTable_wksp (HUF_CElt* tree, const U32* count, U32 maxSymbolValu
huffNode[n].nbBits = huffNode[ huffNode[n].parent ].nbBits + 1;
/* enforce maxTableLog */
maxNbBits = HUF_setMaxHeight(huffNode, nonNullRank, maxNbBits);
maxNbBits = HUF_setMaxHeight(huffNode, (U32)nonNullRank, maxNbBits);
/* fill result into tree (val, nbBits) */
{ U16 nbPerRank[HUF_TABLELOG_MAX+1] = {0};
U16 valPerRank[HUF_TABLELOG_MAX+1] = {0};
int const alphabetSize = (int)(maxSymbolValue + 1);
if (maxNbBits > HUF_TABLELOG_MAX) return ERROR(GENERIC); /* check fit into table */
for (n=0; n<=nonNullRank; n++)
nbPerRank[huffNode[n].nbBits]++;
/* determine stating value per rank */
{ U16 min = 0;
for (n=maxNbBits; n>0; n--) {
for (n=(int)maxNbBits; n>0; n--) {
valPerRank[n] = min; /* get starting value within each rank */
min += nbPerRank[n];
min >>= 1;
} }
for (n=0; n<=maxSymbolValue; n++)
for (n=0; n<alphabetSize; n++)
tree[huffNode[n].byte].nbBits = huffNode[n].nbBits; /* push nbBits per symbol, symbol order */
for (n=0; n<=maxSymbolValue; n++)
for (n=0; n<alphabetSize; n++)
tree[n].val = valPerRank[tree[n].nbBits]++; /* assign value within rank, symbol order */
}
@ -425,22 +413,13 @@ size_t HUF_buildCTable_wksp (HUF_CElt* tree, const U32* count, U32 maxSymbolValu
* @return : maxNbBits
* Note : count is used before tree is written, so they can safely overlap
*/
size_t HUF_buildCTable (HUF_CElt* tree, const U32* count, U32 maxSymbolValue, U32 maxNbBits)
size_t HUF_buildCTable (HUF_CElt* tree, const unsigned* count, unsigned maxSymbolValue, unsigned maxNbBits)
{
huffNodeTable* nodeTable = ExAllocatePoolWithTag(NonPagedPool, sizeof(huffNodeTable), HUFC_ALLOC_TAG);
size_t ret;
if (!nodeTable)
return 0;
ret = HUF_buildCTable_wksp(tree, count, maxSymbolValue, maxNbBits, nodeTable, sizeof(huffNodeTable));
ExFreePool(nodeTable);
return ret;
HUF_buildCTable_wksp_tables workspace;
return HUF_buildCTable_wksp(tree, count, maxSymbolValue, maxNbBits, &workspace, sizeof(workspace));
}
static size_t HUF_estimateCompressedSize(HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue)
size_t HUF_estimateCompressedSize(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue)
{
size_t nbBits = 0;
int s;
@ -450,7 +429,7 @@ static size_t HUF_estimateCompressedSize(HUF_CElt* CTable, const unsigned* count
return nbBits >> 3;
}
static int HUF_validateCTable(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue) {
int HUF_validateCTable(const HUF_CElt* CTable, const unsigned* count, unsigned maxSymbolValue) {
int bad = 0;
int s;
for (s = 0; s <= (int)maxSymbolValue; ++s) {
@ -489,7 +468,7 @@ HUF_compress1X_usingCTable_internal_body(void* dst, size_t dstSize,
/* init */
if (dstSize < 8) return 0; /* not enough space to compress */
{ size_t const initErr = BIT_initCStream(&bitC, op, oend-op);
{ size_t const initErr = BIT_initCStream(&bitC, op, (size_t)(oend-op));
if (HUF_isError(initErr)) return 0; }
n = srcSize & ~3; /* join to mod 4 */
@ -586,7 +565,8 @@ HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize,
if (srcSize < 12) return 0; /* no saving possible : too small input */
op += 6; /* jumpTable */
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, oend-op, ip, segmentSize, CTable, bmi2) );
assert(op <= oend);
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, bmi2) );
if (cSize==0) return 0;
assert(cSize <= 65535);
MEM_writeLE16(ostart, (U16)cSize);
@ -594,7 +574,8 @@ HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize,
}
ip += segmentSize;
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, oend-op, ip, segmentSize, CTable, bmi2) );
assert(op <= oend);
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, bmi2) );
if (cSize==0) return 0;
assert(cSize <= 65535);
MEM_writeLE16(ostart+2, (U16)cSize);
@ -602,7 +583,8 @@ HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize,
}
ip += segmentSize;
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, oend-op, ip, segmentSize, CTable, bmi2) );
assert(op <= oend);
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, segmentSize, CTable, bmi2) );
if (cSize==0) return 0;
assert(cSize <= 65535);
MEM_writeLE16(ostart+4, (U16)cSize);
@ -610,12 +592,14 @@ HUF_compress4X_usingCTable_internal(void* dst, size_t dstSize,
}
ip += segmentSize;
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, oend-op, ip, iend-ip, CTable, bmi2) );
assert(op <= oend);
assert(ip <= iend);
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable_internal(op, (size_t)(oend-op), ip, (size_t)(iend-ip), CTable, bmi2) );
if (cSize==0) return 0;
op += cSize;
}
return op-ostart;
return (size_t)(op-ostart);
}
size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable)
@ -623,36 +607,38 @@ size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, si
return HUF_compress4X_usingCTable_internal(dst, dstSize, src, srcSize, CTable, /* bmi2 */ 0);
}
typedef enum { HUF_singleStream, HUF_fourStreams } HUF_nbStreams_e;
static size_t HUF_compressCTable_internal(
BYTE* const ostart, BYTE* op, BYTE* const oend,
const void* src, size_t srcSize,
unsigned singleStream, const HUF_CElt* CTable, const int bmi2)
HUF_nbStreams_e nbStreams, const HUF_CElt* CTable, const int bmi2)
{
size_t const cSize = singleStream ?
HUF_compress1X_usingCTable_internal(op, oend - op, src, srcSize, CTable, bmi2) :
HUF_compress4X_usingCTable_internal(op, oend - op, src, srcSize, CTable, bmi2);
size_t const cSize = (nbStreams==HUF_singleStream) ?
HUF_compress1X_usingCTable_internal(op, (size_t)(oend - op), src, srcSize, CTable, bmi2) :
HUF_compress4X_usingCTable_internal(op, (size_t)(oend - op), src, srcSize, CTable, bmi2);
if (HUF_isError(cSize)) { return cSize; }
if (cSize==0) { return 0; } /* uncompressible */
op += cSize;
/* check compressibility */
assert(op >= ostart);
if ((size_t)(op-ostart) >= srcSize-1) { return 0; }
return op-ostart;
return (size_t)(op-ostart);
}
typedef struct {
U32 count[HUF_SYMBOLVALUE_MAX + 1];
unsigned count[HUF_SYMBOLVALUE_MAX + 1];
HUF_CElt CTable[HUF_SYMBOLVALUE_MAX + 1];
huffNodeTable nodeTable;
HUF_buildCTable_wksp_tables buildCTable_wksp;
} HUF_compress_tables_t;
/* HUF_compress_internal() :
* `workSpace` must a table of at least HUF_WORKSPACE_SIZE_U32 unsigned */
static size_t HUF_compress_internal (
void* dst, size_t dstSize,
static size_t
HUF_compress_internal (void* dst, size_t dstSize,
const void* src, size_t srcSize,
unsigned maxSymbolValue, unsigned huffLog,
unsigned singleStream,
HUF_nbStreams_e nbStreams,
void* workSpace, size_t wkspSize,
HUF_CElt* oldHufTable, HUF_repeat* repeat, int preferRepeat,
const int bmi2)
@ -662,9 +648,11 @@ static size_t HUF_compress_internal (
BYTE* const oend = ostart + dstSize;
BYTE* op = ostart;
HUF_STATIC_ASSERT(sizeof(*table) <= HUF_WORKSPACE_SIZE);
/* checks & inits */
if (((size_t)workSpace & 3) != 0) return ERROR(GENERIC); /* must be aligned on 4-bytes boundaries */
if (wkspSize < sizeof(*table)) return ERROR(workSpace_tooSmall);
if (wkspSize < HUF_WORKSPACE_SIZE) return ERROR(workSpace_tooSmall);
if (!srcSize) return 0; /* Uncompressed */
if (!dstSize) return 0; /* cannot fit anything within dst budget */
if (srcSize > HUF_BLOCKSIZE_MAX) return ERROR(srcSize_wrong); /* current block size limit */
@ -677,11 +665,11 @@ static size_t HUF_compress_internal (
if (preferRepeat && repeat && *repeat == HUF_repeat_valid) {
return HUF_compressCTable_internal(ostart, op, oend,
src, srcSize,
singleStream, oldHufTable, bmi2);
nbStreams, oldHufTable, bmi2);
}
/* Scan input and build symbol stats */
{ CHECK_V_F(largest, HIST_count_wksp (table->count, &maxSymbolValue, (const BYTE*)src, srcSize, table->count) );
{ CHECK_V_F(largest, HIST_count_wksp (table->count, &maxSymbolValue, (const BYTE*)src, srcSize, workSpace, wkspSize) );
if (largest == srcSize) { *ostart = ((const BYTE*)src)[0]; return 1; } /* single symbol, rle */
if (largest <= (srcSize >> 7)+4) return 0; /* heuristic : probably not compressible enough */
}
@ -696,14 +684,15 @@ static size_t HUF_compress_internal (
if (preferRepeat && repeat && *repeat != HUF_repeat_none) {
return HUF_compressCTable_internal(ostart, op, oend,
src, srcSize,
singleStream, oldHufTable, bmi2);
nbStreams, oldHufTable, bmi2);
}
/* Build Huffman Tree */
huffLog = HUF_optimalTableLog(huffLog, srcSize, maxSymbolValue);
{ CHECK_V_F(maxBits, HUF_buildCTable_wksp(table->CTable, table->count,
{ size_t const maxBits = HUF_buildCTable_wksp(table->CTable, table->count,
maxSymbolValue, huffLog,
table->nodeTable, sizeof(table->nodeTable)) );
&table->buildCTable_wksp, sizeof(table->buildCTable_wksp));
CHECK_F(maxBits);
huffLog = (U32)maxBits;
/* Zero unused symbols in CTable, so we can check it for validity */
memset(table->CTable + (maxSymbolValue + 1), 0,
@ -719,7 +708,7 @@ static size_t HUF_compress_internal (
if (oldSize <= hSize + newSize || hSize + 12 >= srcSize) {
return HUF_compressCTable_internal(ostart, op, oend,
src, srcSize,
singleStream, oldHufTable, bmi2);
nbStreams, oldHufTable, bmi2);
} }
/* Use the new huffman table */
@ -731,7 +720,7 @@ static size_t HUF_compress_internal (
}
return HUF_compressCTable_internal(ostart, op, oend,
src, srcSize,
singleStream, table->CTable, bmi2);
nbStreams, table->CTable, bmi2);
}
@ -741,7 +730,7 @@ size_t HUF_compress1X_wksp (void* dst, size_t dstSize,
void* workSpace, size_t wkspSize)
{
return HUF_compress_internal(dst, dstSize, src, srcSize,
maxSymbolValue, huffLog, 1 /*single stream*/,
maxSymbolValue, huffLog, HUF_singleStream,
workSpace, wkspSize,
NULL, NULL, 0, 0 /*bmi2*/);
}
@ -753,7 +742,7 @@ size_t HUF_compress1X_repeat (void* dst, size_t dstSize,
HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2)
{
return HUF_compress_internal(dst, dstSize, src, srcSize,
maxSymbolValue, huffLog, 1 /*single stream*/,
maxSymbolValue, huffLog, HUF_singleStream,
workSpace, wkspSize, hufTable,
repeat, preferRepeat, bmi2);
}
@ -762,17 +751,8 @@ size_t HUF_compress1X (void* dst, size_t dstSize,
const void* src, size_t srcSize,
unsigned maxSymbolValue, unsigned huffLog)
{
unsigned* workSpace = ExAllocatePoolWithTag(NonPagedPool, sizeof(unsigned) * HUF_WORKSPACE_SIZE_U32, HUFC_ALLOC_TAG);
size_t ret;
if (!workSpace)
return 0;
ret = HUF_compress1X_wksp(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, workSpace, sizeof(unsigned) * HUF_WORKSPACE_SIZE_U32);
ExFreePool(workSpace);
return ret;
unsigned workSpace[HUF_WORKSPACE_SIZE_U32];
return HUF_compress1X_wksp(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, workSpace, sizeof(workSpace));
}
/* HUF_compress4X_repeat():
@ -784,7 +764,7 @@ size_t HUF_compress4X_wksp (void* dst, size_t dstSize,
void* workSpace, size_t wkspSize)
{
return HUF_compress_internal(dst, dstSize, src, srcSize,
maxSymbolValue, huffLog, 0 /*4 streams*/,
maxSymbolValue, huffLog, HUF_fourStreams,
workSpace, wkspSize,
NULL, NULL, 0, 0 /*bmi2*/);
}
@ -799,7 +779,7 @@ size_t HUF_compress4X_repeat (void* dst, size_t dstSize,
HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2)
{
return HUF_compress_internal(dst, dstSize, src, srcSize,
maxSymbolValue, huffLog, 0 /* 4 streams */,
maxSymbolValue, huffLog, HUF_fourStreams,
workSpace, wkspSize,
hufTable, repeat, preferRepeat, bmi2);
}
@ -808,17 +788,8 @@ size_t HUF_compress2 (void* dst, size_t dstSize,
const void* src, size_t srcSize,
unsigned maxSymbolValue, unsigned huffLog)
{
unsigned* workSpace = ExAllocatePoolWithTag(NonPagedPool, sizeof(unsigned) * HUF_WORKSPACE_SIZE_U32, HUFC_ALLOC_TAG);
size_t ret;
if (!workSpace)
return 0;
ret = HUF_compress4X_wksp(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, workSpace, sizeof(unsigned) * HUF_WORKSPACE_SIZE_U32);
ExFreePool(workSpace);
return ret;
unsigned workSpace[HUF_WORKSPACE_SIZE_U32];
return HUF_compress4X_wksp(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, workSpace, sizeof(workSpace));
}
size_t HUF_compress (void* dst, size_t maxDstSize, const void* src, size_t srcSize)

416
drivers/filesystems/btrfs/zstd/huf_decompress.c
View file

@ -1,35 +1,15 @@
/* ******************************************************************
huff0 huffman decoder,
part of Finite State Entropy library
Copyright (C) 2013-present, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy
* huff0 huffman decoder,
* part of Finite State Entropy library
* Copyright (c) 2013-2020, Yann Collet, Facebook, Inc.
*
* You can contact the author at :
* - FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
****************************************************************** */
/* **************************************************************
@ -42,17 +22,25 @@
#define HUF_STATIC_LINKING_ONLY
#include "huf.h"
#include "error_private.h"
#include <ntifs.h>
#include <ntddk.h>
#define HUFD_ALLOC_TAG 0x64465548 // "HUFd"
/* **************************************************************
* Macros
****************************************************************/
/* These two optional macros force the use one way or another of the two
* Huffman decompression implementations. You can't force in both directions
* at the same time.
*/
#if defined(HUF_FORCE_DECOMPRESS_X1) && \
defined(HUF_FORCE_DECOMPRESS_X2)
#error "Cannot force the use of the X1 and X2 decoders at the same time!"
#endif
/* **************************************************************
* Error Management
****************************************************************/
#define HUF_isError ERR_isError
#define CHECK_F(f) { size_t const err_ = (f); if (HUF_isError(err_)) return err_; }
/* **************************************************************
@ -62,6 +50,51 @@
#define HUF_ALIGN_MASK(x, mask) (((x) + (mask)) & ~(mask))
/* **************************************************************
* BMI2 Variant Wrappers
****************************************************************/
#if DYNAMIC_BMI2
#define HUF_DGEN(fn) \
\
static size_t fn##_default( \
void* dst, size_t dstSize, \
const void* cSrc, size_t cSrcSize, \
const HUF_DTable* DTable) \
{ \
return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \
} \
\
static TARGET_ATTRIBUTE("bmi2") size_t fn##_bmi2( \
void* dst, size_t dstSize, \
const void* cSrc, size_t cSrcSize, \
const HUF_DTable* DTable) \
{ \
return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \
} \
\
static size_t fn(void* dst, size_t dstSize, void const* cSrc, \
size_t cSrcSize, HUF_DTable const* DTable, int bmi2) \
{ \
if (bmi2) { \
return fn##_bmi2(dst, dstSize, cSrc, cSrcSize, DTable); \
} \
return fn##_default(dst, dstSize, cSrc, cSrcSize, DTable); \
}
#else
#define HUF_DGEN(fn) \
static size_t fn(void* dst, size_t dstSize, void const* cSrc, \
size_t cSrcSize, HUF_DTable const* DTable, int bmi2) \
{ \
(void)bmi2; \
return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \
}
#endif
/*-***************************/
/* generic DTableDesc */
/*-***************************/
@ -75,6 +108,8 @@ static DTableDesc HUF_getDTableDesc(const HUF_DTable* table)
}
#ifndef HUF_FORCE_DECOMPRESS_X2
/*-***************************/
/* single-symbol decoding */
/*-***************************/
@ -123,17 +158,29 @@ size_t HUF_readDTableX1_wksp(HUF_DTable* DTable, const void* src, size_t srcSize
/* fill DTable */
{ U32 n;
for (n=0; n<nbSymbols; n++) {
U32 const w = huffWeight[n];
U32 const length = (1 << w) >> 1;
U32 u;
size_t const nEnd = nbSymbols;
for (n=0; n<nEnd; n++) {
size_t const w = huffWeight[n];
size_t const length = (1 << w) >> 1;
size_t const uStart = rankVal[w];
size_t const uEnd = uStart + length;
size_t u;
HUF_DEltX1 D;
D.byte = (BYTE)n; D.nbBits = (BYTE)(tableLog + 1 - w);
for (u = rankVal[w]; u < rankVal[w] + length; u++)
dt[u] = D;
rankVal[w] += length;
} }
D.byte = (BYTE)n;
D.nbBits = (BYTE)(tableLog + 1 - w);
rankVal[w] = (U32)uEnd;
if (length < 4) {
/* Use length in the loop bound so the compiler knows it is short. */
for (u = 0; u < length; ++u)
dt[uStart + u] = D;
} else {
/* Unroll the loop 4 times, we know it is a power of 2. */
for (u = uStart; u < uEnd; u += 4) {
dt[u + 0] = D;
dt[u + 1] = D;
dt[u + 2] = D;
dt[u + 3] = D;
} } } }
return iSize;
}
@ -224,6 +271,7 @@ HUF_decompress4X1_usingDTable_internal_body(
{ const BYTE* const istart = (const BYTE*) cSrc;
BYTE* const ostart = (BYTE*) dst;
BYTE* const oend = ostart + dstSize;
BYTE* const olimit = oend - 3;
const void* const dtPtr = DTable + 1;
const HUF_DEltX1* const dt = (const HUF_DEltX1*)dtPtr;
@ -248,9 +296,9 @@ HUF_decompress4X1_usingDTable_internal_body(
BYTE* op2 = opStart2;
BYTE* op3 = opStart3;
BYTE* op4 = opStart4;
U32 endSignal = BIT_DStream_unfinished;
DTableDesc const dtd = HUF_getDTableDesc(DTable);
U32 const dtLog = dtd.tableLog;
U32 endSignal = 1;
if (length4 > cSrcSize) return ERROR(corruption_detected); /* overflow */
CHECK_F( BIT_initDStream(&bitD1, istart1, length1) );
@ -259,8 +307,7 @@ HUF_decompress4X1_usingDTable_internal_body(
CHECK_F( BIT_initDStream(&bitD4, istart4, length4) );
/* up to 16 symbols per loop (4 symbols per stream) in 64-bit mode */
endSignal = BIT_reloadDStream(&bitD1) | BIT_reloadDStream(&bitD2) | BIT_reloadDStream(&bitD3) | BIT_reloadDStream(&bitD4);
while ( (endSignal==BIT_DStream_unfinished) && (op4<(oend-3)) ) {
for ( ; (endSignal) & (op4 < olimit) ; ) {
HUF_DECODE_SYMBOLX1_2(op1, &bitD1);
HUF_DECODE_SYMBOLX1_2(op2, &bitD2);
HUF_DECODE_SYMBOLX1_2(op3, &bitD3);
@ -277,10 +324,10 @@ HUF_decompress4X1_usingDTable_internal_body(
HUF_DECODE_SYMBOLX1_0(op2, &bitD2);
HUF_DECODE_SYMBOLX1_0(op3, &bitD3);
HUF_DECODE_SYMBOLX1_0(op4, &bitD4);
BIT_reloadDStream(&bitD1);
BIT_reloadDStream(&bitD2);
BIT_reloadDStream(&bitD3);
BIT_reloadDStream(&bitD4);
endSignal &= BIT_reloadDStreamFast(&bitD1) == BIT_DStream_unfinished;
endSignal &= BIT_reloadDStreamFast(&bitD2) == BIT_DStream_unfinished;
endSignal &= BIT_reloadDStreamFast(&bitD3) == BIT_DStream_unfinished;
endSignal &= BIT_reloadDStreamFast(&bitD4) == BIT_DStream_unfinished;
}
/* check corruption */
@ -311,46 +358,6 @@ typedef size_t (*HUF_decompress_usingDTable_t)(void *dst, size_t dstSize,
const void *cSrc,
size_t cSrcSize,
const HUF_DTable *DTable);
#if DYNAMIC_BMI2
#define HUF_DGEN(fn) \
\
static size_t fn##_default( \
void* dst, size_t dstSize, \
const void* cSrc, size_t cSrcSize, \
const HUF_DTable* DTable) \
{ \
return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \
} \
\
static TARGET_ATTRIBUTE("bmi2") size_t fn##_bmi2( \
void* dst, size_t dstSize, \
const void* cSrc, size_t cSrcSize, \
const HUF_DTable* DTable) \
{ \
return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \
} \
\
static size_t fn(void* dst, size_t dstSize, void const* cSrc, \
size_t cSrcSize, HUF_DTable const* DTable, int bmi2) \
{ \
if (bmi2) { \
return fn##_bmi2(dst, dstSize, cSrc, cSrcSize, DTable); \
} \
return fn##_default(dst, dstSize, cSrc, cSrcSize, DTable); \
}
#else
#define HUF_DGEN(fn) \
static size_t fn(void* dst, size_t dstSize, void const* cSrc, \
size_t cSrcSize, HUF_DTable const* DTable, int bmi2) \
{ \
(void)bmi2; \
return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \
}
#endif
HUF_DGEN(HUF_decompress1X1_usingDTable_internal)
HUF_DGEN(HUF_decompress4X1_usingDTable_internal)
@ -392,21 +399,8 @@ size_t HUF_decompress1X1_DCtx(HUF_DTable* DCtx, void* dst, size_t dstSize,
size_t HUF_decompress1X1 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_DTable* DTable = ExAllocatePoolWithTag(NonPagedPool, sizeof(HUF_DTable) * HUF_DTABLE_SIZE(HUF_TABLELOG_MAX-1), HUFD_ALLOC_TAG);
size_t ret;
if (!DTable)
return 0;
RtlZeroMemory(DTable, sizeof(HUF_DTable) * HUF_DTABLE_SIZE(HUF_TABLELOG_MAX-1));
DTable[0] = (U32)(HUF_TABLELOG_MAX-1) * 0x01000001;
ret = HUF_decompress1X1_DCtx (DTable, dst, dstSize, cSrc, cSrcSize);
ExFreePool(DTable);
return ret;
HUF_CREATE_STATIC_DTABLEX1(DTable, HUF_TABLELOG_MAX);
return HUF_decompress1X1_DCtx (DTable, dst, dstSize, cSrc, cSrcSize);
}
size_t HUF_decompress4X1_usingDTable(
@ -450,23 +444,14 @@ size_t HUF_decompress4X1_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, cons
}
size_t HUF_decompress4X1 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_DTable* DTable = ExAllocatePoolWithTag(NonPagedPool, sizeof(HUF_DTable) * HUF_DTABLE_SIZE(HUF_TABLELOG_MAX-1), HUFD_ALLOC_TAG);
size_t ret;
if (!DTable)
return 0;
RtlZeroMemory(DTable, sizeof(HUF_DTable) * HUF_DTABLE_SIZE(HUF_TABLELOG_MAX-1));
DTable[0] = (U32)(HUF_TABLELOG_MAX-1) * 0x01000001;
ret = HUF_decompress4X1_DCtx(DTable, dst, dstSize, cSrc, cSrcSize);
ExFreePool(DTable);
return ret;
HUF_CREATE_STATIC_DTABLEX1(DTable, HUF_TABLELOG_MAX);
return HUF_decompress4X1_DCtx(DTable, dst, dstSize, cSrc, cSrcSize);
}
#endif /* HUF_FORCE_DECOMPRESS_X2 */
#ifndef HUF_FORCE_DECOMPRESS_X1
/* *************************/
/* double-symbols decoding */
@ -761,7 +746,6 @@ HUF_decompress1X2_usingDTable_internal_body(
return dstSize;
}
FORCE_INLINE_TEMPLATE size_t
HUF_decompress4X2_usingDTable_internal_body(
void* dst, size_t dstSize,
@ -773,6 +757,7 @@ HUF_decompress4X2_usingDTable_internal_body(
{ const BYTE* const istart = (const BYTE*) cSrc;
BYTE* const ostart = (BYTE*) dst;
BYTE* const oend = ostart + dstSize;
BYTE* const olimit = oend - (sizeof(size_t)-1);
const void* const dtPtr = DTable+1;
const HUF_DEltX2* const dt = (const HUF_DEltX2*)dtPtr;
@ -797,7 +782,7 @@ HUF_decompress4X2_usingDTable_internal_body(
BYTE* op2 = opStart2;
BYTE* op3 = opStart3;
BYTE* op4 = opStart4;
U32 endSignal;
U32 endSignal = 1;
DTableDesc const dtd = HUF_getDTableDesc(DTable);
U32 const dtLog = dtd.tableLog;
@ -808,8 +793,29 @@ HUF_decompress4X2_usingDTable_internal_body(
CHECK_F( BIT_initDStream(&bitD4, istart4, length4) );
/* 16-32 symbols per loop (4-8 symbols per stream) */
endSignal = BIT_reloadDStream(&bitD1) | BIT_reloadDStream(&bitD2) | BIT_reloadDStream(&bitD3) | BIT_reloadDStream(&bitD4);
for ( ; (endSignal==BIT_DStream_unfinished) & (op4<(oend-(sizeof(bitD4.bitContainer)-1))) ; ) {
for ( ; (endSignal) & (op4 < olimit); ) {
#if defined(__clang__) && (defined(__x86_64__) || defined(__i386__))
HUF_DECODE_SYMBOLX2_2(op1, &bitD1);
HUF_DECODE_SYMBOLX2_1(op1, &bitD1);
HUF_DECODE_SYMBOLX2_2(op1, &bitD1);
HUF_DECODE_SYMBOLX2_0(op1, &bitD1);
HUF_DECODE_SYMBOLX2_2(op2, &bitD2);
HUF_DECODE_SYMBOLX2_1(op2, &bitD2);
HUF_DECODE_SYMBOLX2_2(op2, &bitD2);
HUF_DECODE_SYMBOLX2_0(op2, &bitD2);
endSignal &= BIT_reloadDStreamFast(&bitD1) == BIT_DStream_unfinished;
endSignal &= BIT_reloadDStreamFast(&bitD2) == BIT_DStream_unfinished;
HUF_DECODE_SYMBOLX2_2(op3, &bitD3);
HUF_DECODE_SYMBOLX2_1(op3, &bitD3);
HUF_DECODE_SYMBOLX2_2(op3, &bitD3);
HUF_DECODE_SYMBOLX2_0(op3, &bitD3);
HUF_DECODE_SYMBOLX2_2(op4, &bitD4);
HUF_DECODE_SYMBOLX2_1(op4, &bitD4);
HUF_DECODE_SYMBOLX2_2(op4, &bitD4);
HUF_DECODE_SYMBOLX2_0(op4, &bitD4);
endSignal &= BIT_reloadDStreamFast(&bitD3) == BIT_DStream_unfinished;
endSignal &= BIT_reloadDStreamFast(&bitD4) == BIT_DStream_unfinished;
#else
HUF_DECODE_SYMBOLX2_2(op1, &bitD1);
HUF_DECODE_SYMBOLX2_2(op2, &bitD2);
HUF_DECODE_SYMBOLX2_2(op3, &bitD3);
@ -826,8 +832,12 @@ HUF_decompress4X2_usingDTable_internal_body(
HUF_DECODE_SYMBOLX2_0(op2, &bitD2);
HUF_DECODE_SYMBOLX2_0(op3, &bitD3);
HUF_DECODE_SYMBOLX2_0(op4, &bitD4);
endSignal = BIT_reloadDStream(&bitD1) | BIT_reloadDStream(&bitD2) | BIT_reloadDStream(&bitD3) | BIT_reloadDStream(&bitD4);
endSignal = (U32)LIKELY(
(BIT_reloadDStreamFast(&bitD1) == BIT_DStream_unfinished)
& (BIT_reloadDStreamFast(&bitD2) == BIT_DStream_unfinished)
& (BIT_reloadDStreamFast(&bitD3) == BIT_DStream_unfinished)
& (BIT_reloadDStreamFast(&bitD4) == BIT_DStream_unfinished));
#endif
}
/* check corruption */
@ -890,21 +900,8 @@ size_t HUF_decompress1X2_DCtx(HUF_DTable* DCtx, void* dst, size_t dstSize,
size_t HUF_decompress1X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_DTable* DTable = ExAllocatePoolWithTag(NonPagedPool, sizeof(HUF_DTable) * HUF_DTABLE_SIZE(HUF_TABLELOG_MAX), HUFD_ALLOC_TAG);
size_t ret;
if (!DTable)
return 0;
RtlZeroMemory(DTable, sizeof(HUF_DTable) * HUF_DTABLE_SIZE(HUF_TABLELOG_MAX));
DTable[0] = (U32)(HUF_TABLELOG_MAX) * 0x01000001;
ret = HUF_decompress1X2_DCtx(DTable, dst, dstSize, cSrc, cSrcSize);
ExFreePool(DTable);
return ret;
HUF_CREATE_STATIC_DTABLEX2(DTable, HUF_TABLELOG_MAX);
return HUF_decompress1X2_DCtx(DTable, dst, dstSize, cSrc, cSrcSize);
}
size_t HUF_decompress4X2_usingDTable(
@ -950,23 +947,12 @@ size_t HUF_decompress4X2_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize,
size_t HUF_decompress4X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
HUF_DTable* DTable = ExAllocatePoolWithTag(NonPagedPool, sizeof(HUF_DTable) * HUF_DTABLE_SIZE(HUF_TABLELOG_MAX), HUFD_ALLOC_TAG);
size_t ret;
if (!DTable)
return 0;
RtlZeroMemory(DTable, sizeof(HUF_DTable) * HUF_DTABLE_SIZE(HUF_TABLELOG_MAX));
DTable[0] = (U32)(HUF_TABLELOG_MAX) * 0x01000001;
ret = HUF_decompress4X2_DCtx(DTable, dst, dstSize, cSrc, cSrcSize);
ExFreePool(DTable);
return ret;
HUF_CREATE_STATIC_DTABLEX2(DTable, HUF_TABLELOG_MAX);
return HUF_decompress4X2_DCtx(DTable, dst, dstSize, cSrc, cSrcSize);
}
#endif /* HUF_FORCE_DECOMPRESS_X1 */
/* ***********************************/
/* Universal decompression selectors */
@ -977,8 +963,18 @@ size_t HUF_decompress1X_usingDTable(void* dst, size_t maxDstSize,
const HUF_DTable* DTable)
{
DTableDesc const dtd = HUF_getDTableDesc(DTable);
#if defined(HUF_FORCE_DECOMPRESS_X1)
(void)dtd;
assert(dtd.tableType == 0);
return HUF_decompress1X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0);
#elif defined(HUF_FORCE_DECOMPRESS_X2)
(void)dtd;
assert(dtd.tableType == 1);
return HUF_decompress1X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0);
#else
return dtd.tableType ? HUF_decompress1X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0) :
HUF_decompress1X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0);
#endif
}
size_t HUF_decompress4X_usingDTable(void* dst, size_t maxDstSize,
@ -986,11 +982,22 @@ size_t HUF_decompress4X_usingDTable(void* dst, size_t maxDstSize,
const HUF_DTable* DTable)
{
DTableDesc const dtd = HUF_getDTableDesc(DTable);
#if defined(HUF_FORCE_DECOMPRESS_X1)
(void)dtd;
assert(dtd.tableType == 0);
return HUF_decompress4X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0);
#elif defined(HUF_FORCE_DECOMPRESS_X2)
(void)dtd;
assert(dtd.tableType == 1);
return HUF_decompress4X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0);
#else
return dtd.tableType ? HUF_decompress4X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0) :
HUF_decompress4X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0);
#endif
}
#if !defined(HUF_FORCE_DECOMPRESS_X1) && !defined(HUF_FORCE_DECOMPRESS_X2)
typedef struct { U32 tableTime; U32 decode256Time; } algo_time_t;
static const algo_time_t algoTime[16 /* Quantization */][3 /* single, double, quad */] =
{
@ -1012,6 +1019,7 @@ static const algo_time_t algoTime[16 /* Quantization */][3 /* single, double, qu
{{1455,128}, {2422,124}, {4174,124}}, /* Q ==14 : 87-93% */
{{ 722,128}, {1891,145}, {1936,146}}, /* Q ==15 : 93-99% */
};
#endif
/** HUF_selectDecoder() :
* Tells which decoder is likely to decode faster,
@ -1022,6 +1030,15 @@ U32 HUF_selectDecoder (size_t dstSize, size_t cSrcSize)
{
assert(dstSize > 0);
assert(dstSize <= 128*1024);
#if defined(HUF_FORCE_DECOMPRESS_X1)
(void)dstSize;
(void)cSrcSize;
return 0;
#elif defined(HUF_FORCE_DECOMPRESS_X2)
(void)dstSize;
(void)cSrcSize;
return 1;
#else
/* decoder timing evaluation */
{ U32 const Q = (cSrcSize >= dstSize) ? 15 : (U32)(cSrcSize * 16 / dstSize); /* Q < 16 */
U32 const D256 = (U32)(dstSize >> 8);
@ -1029,14 +1046,18 @@ U32 HUF_selectDecoder (size_t dstSize, size_t cSrcSize)
U32 DTime1 = algoTime[Q][1].tableTime + (algoTime[Q][1].decode256Time * D256);
DTime1 += DTime1 >> 3; /* advantage to algorithm using less memory, to reduce cache eviction */
return DTime1 < DTime0;
} }
}
#endif
}
typedef size_t (*decompressionAlgo)(void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize);
size_t HUF_decompress (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize)
{
#if !defined(HUF_FORCE_DECOMPRESS_X1) && !defined(HUF_FORCE_DECOMPRESS_X2)
static const decompressionAlgo decompress[2] = { HUF_decompress4X1, HUF_decompress4X2 };
#endif
/* validation checks */
if (dstSize == 0) return ERROR(dstSize_tooSmall);
@ -1045,7 +1066,17 @@ size_t HUF_decompress (void* dst, size_t dstSize, const void* cSrc, size_t cSrcS
if (cSrcSize == 1) { memset(dst, *(const BYTE*)cSrc, dstSize); return dstSize; } /* RLE */
{ U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize);
#if defined(HUF_FORCE_DECOMPRESS_X1)
(void)algoNb;
assert(algoNb == 0);
return HUF_decompress4X1(dst, dstSize, cSrc, cSrcSize);
#elif defined(HUF_FORCE_DECOMPRESS_X2)
(void)algoNb;
assert(algoNb == 1);
return HUF_decompress4X2(dst, dstSize, cSrc, cSrcSize);
#else
return decompress[algoNb](dst, dstSize, cSrc, cSrcSize);
#endif
}
}
@ -1058,8 +1089,18 @@ size_t HUF_decompress4X_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const
if (cSrcSize == 1) { memset(dst, *(const BYTE*)cSrc, dstSize); return dstSize; } /* RLE */
{ U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize);
#if defined(HUF_FORCE_DECOMPRESS_X1)
(void)algoNb;
assert(algoNb == 0);
return HUF_decompress4X1_DCtx(dctx, dst, dstSize, cSrc, cSrcSize);
#elif defined(HUF_FORCE_DECOMPRESS_X2)
(void)algoNb;
assert(algoNb == 1);
return HUF_decompress4X2_DCtx(dctx, dst, dstSize, cSrc, cSrcSize);
#else
return algoNb ? HUF_decompress4X2_DCtx(dctx, dst, dstSize, cSrc, cSrcSize) :
HUF_decompress4X1_DCtx(dctx, dst, dstSize, cSrc, cSrcSize) ;
#endif
}
}
@ -1081,8 +1122,19 @@ size_t HUF_decompress4X_hufOnly_wksp(HUF_DTable* dctx, void* dst,
if (cSrcSize == 0) return ERROR(corruption_detected);
{ U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize);
return algoNb ? HUF_decompress4X2_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize):
#if defined(HUF_FORCE_DECOMPRESS_X1)
(void)algoNb;
assert(algoNb == 0);
return HUF_decompress4X1_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize);
#elif defined(HUF_FORCE_DECOMPRESS_X2)
(void)algoNb;
assert(algoNb == 1);
return HUF_decompress4X2_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize);
#else
return algoNb ? HUF_decompress4X2_DCtx_wksp(dctx, dst, dstSize, cSrc,
cSrcSize, workSpace, wkspSize):
HUF_decompress4X1_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize);
#endif
}
}
@ -1097,10 +1149,22 @@ size_t HUF_decompress1X_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize,
if (cSrcSize == 1) { memset(dst, *(const BYTE*)cSrc, dstSize); return dstSize; } /* RLE */
{ U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize);
#if defined(HUF_FORCE_DECOMPRESS_X1)
(void)algoNb;
assert(algoNb == 0);
return HUF_decompress1X1_DCtx_wksp(dctx, dst, dstSize, cSrc,
cSrcSize, workSpace, wkspSize);
#elif defined(HUF_FORCE_DECOMPRESS_X2)
(void)algoNb;
assert(algoNb == 1);
return HUF_decompress1X2_DCtx_wksp(dctx, dst, dstSize, cSrc,
cSrcSize, workSpace, wkspSize);
#else
return algoNb ? HUF_decompress1X2_DCtx_wksp(dctx, dst, dstSize, cSrc,
cSrcSize, workSpace, wkspSize):
HUF_decompress1X1_DCtx_wksp(dctx, dst, dstSize, cSrc,
cSrcSize, workSpace, wkspSize);
#endif
}
}
@ -1116,10 +1180,21 @@ size_t HUF_decompress1X_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize,
size_t HUF_decompress1X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2)
{
DTableDesc const dtd = HUF_getDTableDesc(DTable);
#if defined(HUF_FORCE_DECOMPRESS_X1)
(void)dtd;
assert(dtd.tableType == 0);
return HUF_decompress1X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2);
#elif defined(HUF_FORCE_DECOMPRESS_X2)
(void)dtd;
assert(dtd.tableType == 1);
return HUF_decompress1X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2);
#else
return dtd.tableType ? HUF_decompress1X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2) :
HUF_decompress1X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2);
#endif
}
#ifndef HUF_FORCE_DECOMPRESS_X2
size_t HUF_decompress1X1_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2)
{
const BYTE* ip = (const BYTE*) cSrc;
@ -1131,12 +1206,23 @@ size_t HUF_decompress1X1_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstS
return HUF_decompress1X1_usingDTable_internal(dst, dstSize, ip, cSrcSize, dctx, bmi2);
}
#endif
size_t HUF_decompress4X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2)
{
DTableDesc const dtd = HUF_getDTableDesc(DTable);
#if defined(HUF_FORCE_DECOMPRESS_X1)
(void)dtd;
assert(dtd.tableType == 0);
return HUF_decompress4X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2);
#elif defined(HUF_FORCE_DECOMPRESS_X2)
(void)dtd;
assert(dtd.tableType == 1);
return HUF_decompress4X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2);
#else
return dtd.tableType ? HUF_decompress4X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2) :
HUF_decompress4X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2);
#endif
}
size_t HUF_decompress4X_hufOnly_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2)
@ -1146,7 +1232,17 @@ size_t HUF_decompress4X_hufOnly_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t ds
if (cSrcSize == 0) return ERROR(corruption_detected);
{ U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize);
#if defined(HUF_FORCE_DECOMPRESS_X1)
(void)algoNb;
assert(algoNb == 0);
return HUF_decompress4X1_DCtx_wksp_bmi2(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize, bmi2);
#elif defined(HUF_FORCE_DECOMPRESS_X2)
(void)algoNb;
assert(algoNb == 1);
return HUF_decompress4X2_DCtx_wksp_bmi2(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize, bmi2);
#else
return algoNb ? HUF_decompress4X2_DCtx_wksp_bmi2(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize, bmi2) :
HUF_decompress4X1_DCtx_wksp_bmi2(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize, bmi2);
#endif
}
}

77
drivers/filesystems/btrfs/zstd/mem.h
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -47,6 +47,79 @@ extern "C" {
#define MEM_STATIC_ASSERT(c) { enum { MEM_static_assert = 1/(int)(!!(c)) }; }
MEM_STATIC void MEM_check(void) { MEM_STATIC_ASSERT((sizeof(size_t)==4) || (sizeof(size_t)==8)); }
/* detects whether we are being compiled under msan */
#if defined (__has_feature)
# if __has_feature(memory_sanitizer)
# define MEMORY_SANITIZER 1
# endif
#endif
#if defined (MEMORY_SANITIZER)
/* Not all platforms that support msan provide sanitizers/msan_interface.h.
* We therefore declare the functions we need ourselves, rather than trying to
* include the header file... */
#include <stdint.h> /* intptr_t */
/* Make memory region fully initialized (without changing its contents). */
void __msan_unpoison(const volatile void *a, size_t size);
/* Make memory region fully uninitialized (without changing its contents).
This is a legacy interface that does not update origin information. Use
__msan_allocated_memory() instead. */
void __msan_poison(const volatile void *a, size_t size);
/* Returns the offset of the first (at least partially) poisoned byte in the
memory range, or -1 if the whole range is good. */
intptr_t __msan_test_shadow(const volatile void *x, size_t size);
#endif
/* detects whether we are being compiled under asan */
#if defined (__has_feature)
# if __has_feature(address_sanitizer)
# define ADDRESS_SANITIZER 1
# endif
#elif defined(__SANITIZE_ADDRESS__)
# define ADDRESS_SANITIZER 1
#endif
#if defined (ADDRESS_SANITIZER)
/* Not all platforms that support asan provide sanitizers/asan_interface.h.
* We therefore declare the functions we need ourselves, rather than trying to
* include the header file... */
/**
* Marks a memory region (<c>[addr, addr+size)</c>) as unaddressable.
*
* This memory must be previously allocated by your program. Instrumented
* code is forbidden from accessing addresses in this region until it is
* unpoisoned. This function is not guaranteed to poison the entire region -
* it could poison only a subregion of <c>[addr, addr+size)</c> due to ASan
* alignment restrictions.
*
* \note This function is not thread-safe because no two threads can poison or
* unpoison memory in the same memory region simultaneously.
*
* \param addr Start of memory region.
* \param size Size of memory region. */
void __asan_poison_memory_region(void const volatile *addr, size_t size);
/**
* Marks a memory region (<c>[addr, addr+size)</c>) as addressable.
*
* This memory must be previously allocated by your program. Accessing
* addresses in this region is allowed until this region is poisoned again.
* This function could unpoison a super-region of <c>[addr, addr+size)</c> due
* to ASan alignment restrictions.
*
* \note This function is not thread-safe because no two threads can
* poison or unpoison memory in the same memory region simultaneously.
*
* \param addr Start of memory region.
* \param size Size of memory region. */
void __asan_unpoison_memory_region(void const volatile *addr, size_t size);
#endif
/*-**************************************************************
* Basic Types
@ -102,7 +175,7 @@ MEM_STATIC void MEM_check(void) { MEM_STATIC_ASSERT((sizeof(size_t)==4) || (size
#ifndef MEM_FORCE_MEMORY_ACCESS /* can be defined externally, on command line for example */
# if defined(__GNUC__) && ( defined(__ARM_ARCH_6__) || defined(__ARM_ARCH_6J__) || defined(__ARM_ARCH_6K__) || defined(__ARM_ARCH_6Z__) || defined(__ARM_ARCH_6ZK__) || defined(__ARM_ARCH_6T2__) )
# define MEM_FORCE_MEMORY_ACCESS 2
# elif defined(__INTEL_COMPILER) || defined(__GNUC__)
# elif defined(__INTEL_COMPILER) || defined(__GNUC__) || defined(__ICCARM__)
# define MEM_FORCE_MEMORY_ACCESS 1
# endif
#endif

2078
drivers/filesystems/btrfs/zstd/zstd.h
View file

File diff suppressed because it is too large Load diff

6
drivers/filesystems/btrfs/zstd/zstd_common.c
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -30,8 +30,10 @@ const char* ZSTD_versionString(void) { return ZSTD_VERSION_STRING; }
/*-****************************************
* ZSTD Error Management
******************************************/
#undef ZSTD_isError /* defined within zstd_internal.h */
/*! ZSTD_isError() :
* tells if a return value is an error code */
* tells if a return value is an error code
* symbol is required for external callers */
unsigned ZSTD_isError(size_t code) { return ERR_isError(code); }
/*! ZSTD_getErrorName() :

3592
drivers/filesystems/btrfs/zstd/zstd_compress.c
View file

File diff suppressed because it is too large Load diff

499
drivers/filesystems/btrfs/zstd/zstd_compress_internal.h
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -19,6 +19,7 @@
* Dependencies
***************************************/
#include "zstd_internal.h"
#include "zstd_cwksp.h"
#ifdef ZSTD_MULTITHREAD
# include "zstdmt_compress.h"
#endif
@ -33,13 +34,13 @@ extern "C" {
***************************************/
#define kSearchStrength 8
#define HASH_READ_SIZE 8
#define ZSTD_DUBT_UNSORTED_MARK 1 /* For btlazy2 strategy, index 1 now means "unsorted".
#define ZSTD_DUBT_UNSORTED_MARK 1 /* For btlazy2 strategy, index ZSTD_DUBT_UNSORTED_MARK==1 means "unsorted".
It could be confused for a real successor at index "1", if sorted as larger than its predecessor.
It's not a big deal though : candidate will just be sorted again.
Additionnally, candidate position 1 will be lost.
Additionally, candidate position 1 will be lost.
But candidate 1 cannot hide a large tree of candidates, so it's a minimal loss.
The benefit is that ZSTD_DUBT_UNSORTED_MARK cannot be misdhandled after table re-use with a different strategy
Constant required by ZSTD_compressBlock_btlazy2() and ZSTD_reduceTable_internal() */
The benefit is that ZSTD_DUBT_UNSORTED_MARK cannot be mishandled after table re-use with a different strategy.
This constant is required by ZSTD_compressBlock_btlazy2() and ZSTD_reduceTable_internal() */
/*-*************************************
@ -48,18 +49,20 @@ extern "C" {
typedef enum { ZSTDcs_created=0, ZSTDcs_init, ZSTDcs_ongoing, ZSTDcs_ending } ZSTD_compressionStage_e;
typedef enum { zcss_init=0, zcss_load, zcss_flush } ZSTD_cStreamStage;
typedef enum {
ZSTD_dictDefaultAttach = 0,
ZSTD_dictForceAttach = 1,
ZSTD_dictForceCopy = -1,
} ZSTD_dictAttachPref_e;
typedef struct ZSTD_prefixDict_s {
const void* dict;
size_t dictSize;
ZSTD_dictContentType_e dictContentType;
} ZSTD_prefixDict;
typedef struct {
void* dictBuffer;
void const* dict;
size_t dictSize;
ZSTD_dictContentType_e dictContentType;
ZSTD_CDict* cdict;
} ZSTD_localDict;
typedef struct {
U32 CTable[HUF_CTABLE_SIZE_U32(255)];
HUF_repeat repeatMode;
@ -96,10 +99,10 @@ typedef enum { zop_dynamic=0, zop_predef } ZSTD_OptPrice_e;
typedef struct {
/* All tables are allocated inside cctx->workspace by ZSTD_resetCCtx_internal() */
U32* litFreq; /* table of literals statistics, of size 256 */
U32* litLengthFreq; /* table of litLength statistics, of size (MaxLL+1) */
U32* matchLengthFreq; /* table of matchLength statistics, of size (MaxML+1) */
U32* offCodeFreq; /* table of offCode statistics, of size (MaxOff+1) */
unsigned* litFreq; /* table of literals statistics, of size 256 */
unsigned* litLengthFreq; /* table of litLength statistics, of size (MaxLL+1) */
unsigned* matchLengthFreq; /* table of matchLength statistics, of size (MaxML+1) */
unsigned* offCodeFreq; /* table of offCode statistics, of size (MaxOff+1) */
ZSTD_match_t* matchTable; /* list of found matches, of size ZSTD_OPT_NUM+1 */
ZSTD_optimal_t* priceTable; /* All positions tracked by optimal parser, of size ZSTD_OPT_NUM+1 */
@ -113,6 +116,7 @@ typedef struct {
U32 offCodeSumBasePrice; /* to compare to log2(offreq) */
ZSTD_OptPrice_e priceType; /* prices can be determined dynamically, or follow a pre-defined cost structure */
const ZSTD_entropyCTables_t* symbolCosts; /* pre-calculated dictionary statistics */
ZSTD_literalCompressionMode_e literalCompressionMode;
} optState_t;
typedef struct {
@ -125,21 +129,26 @@ typedef struct {
BYTE const* base; /* All regular indexes relative to this position */
BYTE const* dictBase; /* extDict indexes relative to this position */
U32 dictLimit; /* below that point, need extDict */
U32 lowLimit; /* below that point, no more data */
U32 lowLimit; /* below that point, no more valid data */
} ZSTD_window_t;
typedef struct ZSTD_matchState_t ZSTD_matchState_t;
struct ZSTD_matchState_t {
ZSTD_window_t window; /* State for window round buffer management */
U32 loadedDictEnd; /* index of end of dictionary */
U32 loadedDictEnd; /* index of end of dictionary, within context's referential.
* When loadedDictEnd != 0, a dictionary is in use, and still valid.
* This relies on a mechanism to set loadedDictEnd=0 when dictionary is no longer within distance.
* Such mechanism is provided within ZSTD_window_enforceMaxDist() and ZSTD_checkDictValidity().
* When dict referential is copied into active context (i.e. not attached),
* loadedDictEnd == dictSize, since referential starts from zero.
*/
U32 nextToUpdate; /* index from which to continue table update */
U32 nextToUpdate3; /* index from which to continue table update */
U32 hashLog3; /* dispatch table : larger == faster, more memory */
U32 hashLog3; /* dispatch table for matches of len==3 : larger == faster, more memory */
U32* hashTable;
U32* hashTable3;
U32* chainTable;
optState_t opt; /* optimal parser state */
const ZSTD_matchState_t *dictMatchState;
const ZSTD_matchState_t* dictMatchState;
ZSTD_compressionParameters cParams;
};
@ -157,6 +166,7 @@ typedef struct {
typedef struct {
ZSTD_window_t window; /* State for the window round buffer management */
ldmEntry_t* hashTable;
U32 loadedDictEnd;
BYTE* bucketOffsets; /* Next position in bucket to insert entry */
U64 hashPower; /* Used to compute the rolling hash.
* Depends on ldmParams.minMatchLength */
@ -167,7 +177,7 @@ typedef struct {
U32 hashLog; /* Log size of hashTable */
U32 bucketSizeLog; /* Log bucket size for collision resolution, at most 8 */
U32 minMatchLength; /* Minimum match length */
U32 hashEveryLog; /* Log number of entries to skip */
U32 hashRateLog; /* Log number of entries to skip */
U32 windowLog; /* Window log for the LDM */
} ldmParams_t;
@ -184,6 +194,13 @@ typedef struct {
size_t capacity; /* The capacity starting from `seq` pointer */
} rawSeqStore_t;
typedef struct {
int collectSequences;
ZSTD_Sequence* seqStart;
size_t seqIndex;
size_t maxSequences;
} SeqCollector;
struct ZSTD_CCtx_params_s {
ZSTD_format_e format;
ZSTD_compressionParameters cParams;
@ -192,13 +209,21 @@ struct ZSTD_CCtx_params_s {
int compressionLevel;
int forceWindow; /* force back-references to respect limit of
* 1<<wLog, even for dictionary */
size_t targetCBlockSize; /* Tries to fit compressed block size to be around targetCBlockSize.
* No target when targetCBlockSize == 0.
* There is no guarantee on compressed block size */
int srcSizeHint; /* User's best guess of source size.
* Hint is not valid when srcSizeHint == 0.
* There is no guarantee that hint is close to actual source size */
ZSTD_dictAttachPref_e attachDictPref;
ZSTD_literalCompressionMode_e literalCompressionMode;
/* Multithreading: used to pass parameters to mtctx */
unsigned nbWorkers;
unsigned jobSize;
unsigned overlapSizeLog;
int nbWorkers;
size_t jobSize;
int overlapLog;
int rsyncable;
/* Long distance matching parameters */
ldmParams_t ldmParams;
@ -215,9 +240,7 @@ struct ZSTD_CCtx_s {
ZSTD_CCtx_params appliedParams;
U32 dictID;
int workSpaceOversizedDuration;
void* workSpace;
size_t workSpaceSize;
ZSTD_cwksp workspace; /* manages buffer for dynamic allocations */
size_t blockSize;
unsigned long long pledgedSrcSizePlusOne; /* this way, 0 (default) == unknown */
unsigned long long consumedSrcSize;
@ -225,6 +248,9 @@ struct ZSTD_CCtx_s {
XXH64_state_t xxhState;
ZSTD_customMem customMem;
size_t staticSize;
SeqCollector seqCollector;
int isFirstBlock;
int initialized;
seqStore_t seqStore; /* sequences storage ptrs */
ldmState_t ldmState; /* long distance matching state */
@ -248,7 +274,7 @@ struct ZSTD_CCtx_s {
U32 frameEnded;
/* Dictionary */
ZSTD_CDict* cdictLocal;
ZSTD_localDict localDict;
const ZSTD_CDict* cdict;
ZSTD_prefixDict prefixDict; /* single-usage dictionary */
@ -300,26 +326,145 @@ MEM_STATIC U32 ZSTD_MLcode(U32 mlBase)
return (mlBase > 127) ? ZSTD_highbit32(mlBase) + ML_deltaCode : ML_Code[mlBase];
}
/*! ZSTD_storeSeq() :
* Store a sequence (literal length, literals, offset code and match length code) into seqStore_t.
* `offsetCode` : distance to match + 3 (values 1-3 are repCodes).
* `mlBase` : matchLength - MINMATCH
*/
MEM_STATIC void ZSTD_storeSeq(seqStore_t* seqStorePtr, size_t litLength, const void* literals, U32 offsetCode, size_t mlBase)
typedef struct repcodes_s {
U32 rep[3];
} repcodes_t;
MEM_STATIC repcodes_t ZSTD_updateRep(U32 const rep[3], U32 const offset, U32 const ll0)
{
repcodes_t newReps;
if (offset >= ZSTD_REP_NUM) { /* full offset */
newReps.rep[2] = rep[1];
newReps.rep[1] = rep[0];
newReps.rep[0] = offset - ZSTD_REP_MOVE;
} else { /* repcode */
U32 const repCode = offset + ll0;
if (repCode > 0) { /* note : if repCode==0, no change */
U32 const currentOffset = (repCode==ZSTD_REP_NUM) ? (rep[0] - 1) : rep[repCode];
newReps.rep[2] = (repCode >= 2) ? rep[1] : rep[2];
newReps.rep[1] = rep[0];
newReps.rep[0] = currentOffset;
} else { /* repCode == 0 */
memcpy(&newReps, rep, sizeof(newReps));
}
}
return newReps;
}
/* ZSTD_cParam_withinBounds:
* @return 1 if value is within cParam bounds,
* 0 otherwise */
MEM_STATIC int ZSTD_cParam_withinBounds(ZSTD_cParameter cParam, int value)
{
ZSTD_bounds const bounds = ZSTD_cParam_getBounds(cParam);
if (ZSTD_isError(bounds.error)) return 0;
if (value < bounds.lowerBound) return 0;
if (value > bounds.upperBound) return 0;
return 1;
}
/* ZSTD_noCompressBlock() :
* Writes uncompressed block to dst buffer from given src.
* Returns the size of the block */
MEM_STATIC size_t ZSTD_noCompressBlock (void* dst, size_t dstCapacity, const void* src, size_t srcSize, U32 lastBlock)
{
U32 const cBlockHeader24 = lastBlock + (((U32)bt_raw)<<1) + (U32)(srcSize << 3);
RETURN_ERROR_IF(srcSize + ZSTD_blockHeaderSize > dstCapacity,
dstSize_tooSmall, "dst buf too small for uncompressed block");
MEM_writeLE24(dst, cBlockHeader24);
memcpy((BYTE*)dst + ZSTD_blockHeaderSize, src, srcSize);
return ZSTD_blockHeaderSize + srcSize;
}
MEM_STATIC size_t ZSTD_rleCompressBlock (void* dst, size_t dstCapacity, BYTE src, size_t srcSize, U32 lastBlock)
{
BYTE* const op = (BYTE*)dst;
U32 const cBlockHeader = lastBlock + (((U32)bt_rle)<<1) + (U32)(srcSize << 3);
RETURN_ERROR_IF(dstCapacity < 4, dstSize_tooSmall, "");
MEM_writeLE24(op, cBlockHeader);
op[3] = src;
return 4;
}
/* ZSTD_minGain() :
* minimum compression required
* to generate a compress block or a compressed literals section.
* note : use same formula for both situations */
MEM_STATIC size_t ZSTD_minGain(size_t srcSize, ZSTD_strategy strat)
{
U32 const minlog = (strat>=ZSTD_btultra) ? (U32)(strat) - 1 : 6;
ZSTD_STATIC_ASSERT(ZSTD_btultra == 8);
assert(ZSTD_cParam_withinBounds(ZSTD_c_strategy, strat));
return (srcSize >> minlog) + 2;
}
MEM_STATIC int ZSTD_disableLiteralsCompression(const ZSTD_CCtx_params* cctxParams)
{
switch (cctxParams->literalCompressionMode) {
case ZSTD_lcm_huffman:
return 0;
case ZSTD_lcm_uncompressed:
return 1;
default:
assert(0 /* impossible: pre-validated */);
/* fall-through */
case ZSTD_lcm_auto:
return (cctxParams->cParams.strategy == ZSTD_fast) && (cctxParams->cParams.targetLength > 0);
}
}
/*! ZSTD_safecopyLiterals() :
* memcpy() function that won't read beyond more than WILDCOPY_OVERLENGTH bytes past ilimit_w.
* Only called when the sequence ends past ilimit_w, so it only needs to be optimized for single
* large copies.
*/
static void ZSTD_safecopyLiterals(BYTE* op, BYTE const* ip, BYTE const* const iend, BYTE const* ilimit_w) {
assert(iend > ilimit_w);
if (ip <= ilimit_w) {
ZSTD_wildcopy(op, ip, ilimit_w - ip, ZSTD_no_overlap);
op += ilimit_w - ip;
ip = ilimit_w;
}
while (ip < iend) *op++ = *ip++;
}
/*! ZSTD_storeSeq() :
* Store a sequence (litlen, litPtr, offCode and mlBase) into seqStore_t.
* `offCode` : distance to match + ZSTD_REP_MOVE (values <= ZSTD_REP_MOVE are repCodes).
* `mlBase` : matchLength - MINMATCH
* Allowed to overread literals up to litLimit.
*/
HINT_INLINE UNUSED_ATTR
void ZSTD_storeSeq(seqStore_t* seqStorePtr, size_t litLength, const BYTE* literals, const BYTE* litLimit, U32 offCode, size_t mlBase)
{
BYTE const* const litLimit_w = litLimit - WILDCOPY_OVERLENGTH;
BYTE const* const litEnd = literals + litLength;
#if defined(DEBUGLEVEL) && (DEBUGLEVEL >= 6)
static const BYTE* g_start = NULL;
if (g_start==NULL) g_start = (const BYTE*)literals; /* note : index only works for compression within a single segment */
{ U32 const pos = (U32)((const BYTE*)literals - g_start);
DEBUGLOG(6, "Cpos%7u :%3u literals, match%4u bytes at offCode%7u",
pos, (U32)litLength, (U32)mlBase+MINMATCH, (U32)offsetCode);
pos, (U32)litLength, (U32)mlBase+MINMATCH, (U32)offCode);
}
#endif
assert((size_t)(seqStorePtr->sequences - seqStorePtr->sequencesStart) < seqStorePtr->maxNbSeq);
/* copy Literals */
assert(seqStorePtr->maxNbLit <= 128 KB);
assert(seqStorePtr->lit + litLength <= seqStorePtr->litStart + seqStorePtr->maxNbLit);
ZSTD_wildcopy(seqStorePtr->lit, literals, litLength);
assert(literals + litLength <= litLimit);
if (litEnd <= litLimit_w) {
/* Common case we can use wildcopy.
* First copy 16 bytes, because literals are likely short.
*/
assert(WILDCOPY_OVERLENGTH >= 16);
ZSTD_copy16(seqStorePtr->lit, literals);
if (litLength > 16) {
ZSTD_wildcopy(seqStorePtr->lit+16, literals+16, (ptrdiff_t)litLength-16, ZSTD_no_overlap);
}
} else {
ZSTD_safecopyLiterals(seqStorePtr->lit, literals, litEnd, litLimit_w);
}
seqStorePtr->lit += litLength;
/* literal Length */
@ -331,7 +476,7 @@ MEM_STATIC void ZSTD_storeSeq(seqStore_t* seqStorePtr, size_t litLength, const v
seqStorePtr->sequences[0].litLength = (U16)litLength;
/* match offset */
seqStorePtr->sequences[0].offset = offsetCode + 1;
seqStorePtr->sequences[0].offset = offCode + 1;
/* match Length */
if (mlBase>0xFFFF) {
@ -354,8 +499,7 @@ static unsigned ZSTD_NbCommonBytes (size_t val)
if (MEM_64bits()) {
# if defined(_MSC_VER) && defined(_WIN64)
unsigned long r = 0;
_BitScanForward64( &r, (U64)val );
return (unsigned)(r>>3);
return _BitScanForward64( &r, (U64)val ) ? (unsigned)(r >> 3) : 0;
# elif defined(__GNUC__) && (__GNUC__ >= 4)
return (__builtin_ctzll((U64)val) >> 3);
# else
@ -372,8 +516,7 @@ static unsigned ZSTD_NbCommonBytes (size_t val)
} else { /* 32 bits */
# if defined(_MSC_VER)
unsigned long r=0;
_BitScanForward( &r, (U32)val );
return (unsigned)(r>>3);
return _BitScanForward( &r, (U32)val ) ? (unsigned)(r >> 3) : 0;
# elif defined(__GNUC__) && (__GNUC__ >= 3)
return (__builtin_ctz((U32)val) >> 3);
# else
@ -388,8 +531,7 @@ static unsigned ZSTD_NbCommonBytes (size_t val)
if (MEM_64bits()) {
# if defined(_MSC_VER) && defined(_WIN64)
unsigned long r = 0;
_BitScanReverse64( &r, val );
return (unsigned)(r>>3);
return _BitScanReverse64( &r, val ) ? (unsigned)(r >> 3) : 0;
# elif defined(__GNUC__) && (__GNUC__ >= 4)
return (__builtin_clzll(val) >> 3);
# else
@ -403,8 +545,7 @@ static unsigned ZSTD_NbCommonBytes (size_t val)
} else { /* 32 bits */
# if defined(_MSC_VER)
unsigned long r = 0;
_BitScanReverse( &r, (unsigned long)val );
return (unsigned)(r>>3);
return _BitScanReverse( &r, (unsigned long)val ) ? (unsigned)(r >> 3) : 0;
# elif defined(__GNUC__) && (__GNUC__ >= 3)
return (__builtin_clz((U32)val) >> 3);
# else
@ -498,9 +639,70 @@ MEM_STATIC size_t ZSTD_hashPtr(const void* p, U32 hBits, U32 mls)
}
}
/** ZSTD_ipow() :
* Return base^exponent.
*/
static U64 ZSTD_ipow(U64 base, U64 exponent)
{
U64 power = 1;
while (exponent) {
if (exponent & 1) power *= base;
exponent >>= 1;
base *= base;
}
return power;
}
#define ZSTD_ROLL_HASH_CHAR_OFFSET 10
/** ZSTD_rollingHash_append() :
* Add the buffer to the hash value.
*/
static U64 ZSTD_rollingHash_append(U64 hash, void const* buf, size_t size)
{
BYTE const* istart = (BYTE const*)buf;
size_t pos;
for (pos = 0; pos < size; ++pos) {
hash *= prime8bytes;
hash += istart[pos] + ZSTD_ROLL_HASH_CHAR_OFFSET;
}
return hash;
}
/** ZSTD_rollingHash_compute() :
* Compute the rolling hash value of the buffer.
*/
MEM_STATIC U64 ZSTD_rollingHash_compute(void const* buf, size_t size)
{
return ZSTD_rollingHash_append(0, buf, size);
}
/** ZSTD_rollingHash_primePower() :
* Compute the primePower to be passed to ZSTD_rollingHash_rotate() for a hash
* over a window of length bytes.
*/
MEM_STATIC U64 ZSTD_rollingHash_primePower(U32 length)
{
return ZSTD_ipow(prime8bytes, length - 1);
}
/** ZSTD_rollingHash_rotate() :
* Rotate the rolling hash by one byte.
*/
MEM_STATIC U64 ZSTD_rollingHash_rotate(U64 hash, BYTE toRemove, BYTE toAdd, U64 primePower)
{
hash -= (toRemove + ZSTD_ROLL_HASH_CHAR_OFFSET) * primePower;
hash *= prime8bytes;
hash += toAdd + ZSTD_ROLL_HASH_CHAR_OFFSET;
return hash;
}
/*-*************************************
* Round buffer management
***************************************/
#if (ZSTD_WINDOWLOG_MAX_64 > 31)
# error "ZSTD_WINDOWLOG_MAX is too large : would overflow ZSTD_CURRENT_MAX"
#endif
/* Max current allowed */
#define ZSTD_CURRENT_MAX ((3U << 29) + (1U << ZSTD_WINDOWLOG_MAX))
/* Maximum chunk size before overflow correction needs to be called again */
@ -590,7 +792,10 @@ MEM_STATIC U32 ZSTD_window_correctOverflow(ZSTD_window_t* window, U32 cycleLog,
*/
U32 const cycleMask = (1U << cycleLog) - 1;
U32 const current = (U32)((BYTE const*)src - window->base);
U32 const newCurrent = (current & cycleMask) + maxDist;
U32 const currentCycle0 = current & cycleMask;
/* Exclude zero so that newCurrent - maxDist >= 1. */
U32 const currentCycle1 = currentCycle0 == 0 ? (1U << cycleLog) : currentCycle0;
U32 const newCurrent = currentCycle1 + maxDist;
U32 const correction = current - newCurrent;
assert((maxDist & cycleMask) == 0);
assert(current > newCurrent);
@ -599,8 +804,17 @@ MEM_STATIC U32 ZSTD_window_correctOverflow(ZSTD_window_t* window, U32 cycleLog,
window->base += correction;
window->dictBase += correction;
window->lowLimit -= correction;
window->dictLimit -= correction;
if (window->lowLimit <= correction) window->lowLimit = 1;
else window->lowLimit -= correction;
if (window->dictLimit <= correction) window->dictLimit = 1;
else window->dictLimit -= correction;
/* Ensure we can still reference the full window. */
assert(newCurrent >= maxDist);
assert(newCurrent - maxDist >= 1);
/* Ensure that lowLimit and dictLimit didn't underflow. */
assert(window->lowLimit <= newCurrent);
assert(window->dictLimit <= newCurrent);
DEBUGLOG(4, "Correction of 0x%x bytes to lowLimit=0x%x", correction,
window->lowLimit);
@ -612,43 +826,107 @@ MEM_STATIC U32 ZSTD_window_correctOverflow(ZSTD_window_t* window, U32 cycleLog,
* Updates lowLimit so that:
* (srcEnd - base) - lowLimit == maxDist + loadedDictEnd
*
* This allows a simple check that index >= lowLimit to see if index is valid.
* This must be called before a block compression call, with srcEnd as the block
* source end.
* It ensures index is valid as long as index >= lowLimit.
* This must be called before a block compression call.
*
* If loadedDictEndPtr is not NULL, we set it to zero once we update lowLimit.
* This is because dictionaries are allowed to be referenced as long as the last
* byte of the dictionary is in the window, but once they are out of range,
* they cannot be referenced. If loadedDictEndPtr is NULL, we use
* loadedDictEnd == 0.
* loadedDictEnd is only defined if a dictionary is in use for current compression.
* As the name implies, loadedDictEnd represents the index at end of dictionary.
* The value lies within context's referential, it can be directly compared to blockEndIdx.
*
* In normal dict mode, the dict is between lowLimit and dictLimit. In
* dictMatchState mode, lowLimit and dictLimit are the same, and the dictionary
* is below them. forceWindow and dictMatchState are therefore incompatible.
* If loadedDictEndPtr is NULL, no dictionary is in use, and we use loadedDictEnd == 0.
* If loadedDictEndPtr is not NULL, we set it to zero after updating lowLimit.
* This is because dictionaries are allowed to be referenced fully
* as long as the last byte of the dictionary is in the window.
* Once input has progressed beyond window size, dictionary cannot be referenced anymore.
*
* In normal dict mode, the dictionary lies between lowLimit and dictLimit.
* In dictMatchState mode, lowLimit and dictLimit are the same,
* and the dictionary is below them.
* forceWindow and dictMatchState are therefore incompatible.
*/
MEM_STATIC void ZSTD_window_enforceMaxDist(ZSTD_window_t* window,
void const* srcEnd, U32 maxDist,
MEM_STATIC void
ZSTD_window_enforceMaxDist(ZSTD_window_t* window,
const void* blockEnd,
U32 maxDist,
U32* loadedDictEndPtr,
const ZSTD_matchState_t** dictMatchStatePtr)
{
U32 const current = (U32)((BYTE const*)srcEnd - window->base);
U32 loadedDictEnd = loadedDictEndPtr != NULL ? *loadedDictEndPtr : 0;
DEBUGLOG(5, "ZSTD_window_enforceMaxDist: current=%u, maxDist=%u", current, maxDist);
if (current > maxDist + loadedDictEnd) {
U32 const newLowLimit = current - maxDist;
U32 const blockEndIdx = (U32)((BYTE const*)blockEnd - window->base);
U32 const loadedDictEnd = (loadedDictEndPtr != NULL) ? *loadedDictEndPtr : 0;
DEBUGLOG(5, "ZSTD_window_enforceMaxDist: blockEndIdx=%u, maxDist=%u, loadedDictEnd=%u",
(unsigned)blockEndIdx, (unsigned)maxDist, (unsigned)loadedDictEnd);
/* - When there is no dictionary : loadedDictEnd == 0.
In which case, the test (blockEndIdx > maxDist) is merely to avoid
overflowing next operation `newLowLimit = blockEndIdx - maxDist`.
- When there is a standard dictionary :
Index referential is copied from the dictionary,
which means it starts from 0.
In which case, loadedDictEnd == dictSize,
and it makes sense to compare `blockEndIdx > maxDist + dictSize`
since `blockEndIdx` also starts from zero.
- When there is an attached dictionary :
loadedDictEnd is expressed within the referential of the context,
so it can be directly compared against blockEndIdx.
*/
if (blockEndIdx > maxDist + loadedDictEnd) {
U32 const newLowLimit = blockEndIdx - maxDist;
if (window->lowLimit < newLowLimit) window->lowLimit = newLowLimit;
if (window->dictLimit < window->lowLimit) {
DEBUGLOG(5, "Update dictLimit to match lowLimit, from %u to %u",
window->dictLimit, window->lowLimit);
(unsigned)window->dictLimit, (unsigned)window->lowLimit);
window->dictLimit = window->lowLimit;
}
if (loadedDictEndPtr)
*loadedDictEndPtr = 0;
if (dictMatchStatePtr)
*dictMatchStatePtr = NULL;
/* On reaching window size, dictionaries are invalidated */
if (loadedDictEndPtr) *loadedDictEndPtr = 0;
if (dictMatchStatePtr) *dictMatchStatePtr = NULL;
}
}
/* Similar to ZSTD_window_enforceMaxDist(),
* but only invalidates dictionary
* when input progresses beyond window size.
* assumption : loadedDictEndPtr and dictMatchStatePtr are valid (non NULL)
* loadedDictEnd uses same referential as window->base
* maxDist is the window size */
MEM_STATIC void
ZSTD_checkDictValidity(const ZSTD_window_t* window,
const void* blockEnd,
U32 maxDist,
U32* loadedDictEndPtr,
const ZSTD_matchState_t** dictMatchStatePtr)
{
assert(loadedDictEndPtr != NULL);
assert(dictMatchStatePtr != NULL);
{ U32 const blockEndIdx = (U32)((BYTE const*)blockEnd - window->base);
U32 const loadedDictEnd = *loadedDictEndPtr;
DEBUGLOG(5, "ZSTD_checkDictValidity: blockEndIdx=%u, maxDist=%u, loadedDictEnd=%u",
(unsigned)blockEndIdx, (unsigned)maxDist, (unsigned)loadedDictEnd);
assert(blockEndIdx >= loadedDictEnd);
if (blockEndIdx > loadedDictEnd + maxDist) {
/* On reaching window size, dictionaries are invalidated.
* For simplification, if window size is reached anywhere within next block,
* the dictionary is invalidated for the full block.
*/
DEBUGLOG(6, "invalidating dictionary for current block (distance > windowSize)");
*loadedDictEndPtr = 0;
*dictMatchStatePtr = NULL;
} else {
if (*loadedDictEndPtr != 0) {
DEBUGLOG(6, "dictionary considered valid for current block");
} } }
}
MEM_STATIC void ZSTD_window_init(ZSTD_window_t* window) {
memset(window, 0, sizeof(*window));
window->base = (BYTE const*)"";
window->dictBase = (BYTE const*)"";
window->dictLimit = 1; /* start from 1, so that 1st position is valid */
window->lowLimit = 1; /* it ensures first and later CCtx usages compress the same */
window->nextSrc = window->base + 1; /* see issue #1241 */
}
/**
* ZSTD_window_update():
* Updates the window by appending [src, src + srcSize) to the window.
@ -662,6 +940,10 @@ MEM_STATIC U32 ZSTD_window_update(ZSTD_window_t* window,
BYTE const* const ip = (BYTE const*)src;
U32 contiguous = 1;
DEBUGLOG(5, "ZSTD_window_update");
if (srcSize == 0)
return contiguous;
assert(window->base != NULL);
assert(window->dictBase != NULL);
/* Check if blocks follow each other */
if (src != window->nextSrc) {
/* not contiguous */
@ -672,7 +954,7 @@ MEM_STATIC U32 ZSTD_window_update(ZSTD_window_t* window,
window->dictLimit = (U32)distanceFromBase;
window->dictBase = window->base;
window->base = ip - distanceFromBase;
// ms->nextToUpdate = window->dictLimit;
/* ms->nextToUpdate = window->dictLimit; */
if (window->dictLimit - window->lowLimit < HASH_READ_SIZE) window->lowLimit = window->dictLimit; /* too small extDict */
contiguous = 0;
}
@ -688,22 +970,52 @@ MEM_STATIC U32 ZSTD_window_update(ZSTD_window_t* window,
return contiguous;
}
/**
* Returns the lowest allowed match index. It may either be in the ext-dict or the prefix.
*/
MEM_STATIC U32 ZSTD_getLowestMatchIndex(const ZSTD_matchState_t* ms, U32 current, unsigned windowLog)
{
U32 const maxDistance = 1U << windowLog;
U32 const lowestValid = ms->window.lowLimit;
U32 const withinWindow = (current - lowestValid > maxDistance) ? current - maxDistance : lowestValid;
U32 const isDictionary = (ms->loadedDictEnd != 0);
U32 const matchLowest = isDictionary ? lowestValid : withinWindow;
return matchLowest;
}
/**
* Returns the lowest allowed match index in the prefix.
*/
MEM_STATIC U32 ZSTD_getLowestPrefixIndex(const ZSTD_matchState_t* ms, U32 current, unsigned windowLog)
{
U32 const maxDistance = 1U << windowLog;
U32 const lowestValid = ms->window.dictLimit;
U32 const withinWindow = (current - lowestValid > maxDistance) ? current - maxDistance : lowestValid;
U32 const isDictionary = (ms->loadedDictEnd != 0);
U32 const matchLowest = isDictionary ? lowestValid : withinWindow;
return matchLowest;
}
/* debug functions */
#if (DEBUGLEVEL>=2)
MEM_STATIC double ZSTD_fWeight(U32 rawStat)
{
U32 const fp_accuracy = 8;
U32 const fp_multiplier = (1 << fp_accuracy);
U32 const stat = rawStat + 1;
U32 const hb = ZSTD_highbit32(stat);
U32 const newStat = rawStat + 1;
U32 const hb = ZSTD_highbit32(newStat);
U32 const BWeight = hb * fp_multiplier;
U32 const FWeight = (stat << fp_accuracy) >> hb;
U32 const FWeight = (newStat << fp_accuracy) >> hb;
U32 const weight = BWeight + FWeight;
assert(hb + fp_accuracy < 31);
return (double)weight / fp_multiplier;
}
/* display a table content,
* listing each element, its frequency, and its predicted bit cost */
MEM_STATIC void ZSTD_debugTable(const U32* table, U32 max)
{
unsigned u, sum;
@ -715,10 +1027,28 @@ MEM_STATIC void ZSTD_debugTable(const U32* table, U32 max)
}
}
#endif
#if defined (__cplusplus)
}
#endif
/* ===============================================================
* Shared internal declarations
* These prototypes may be called from sources not in lib/compress
* =============================================================== */
/* ZSTD_loadCEntropy() :
* dict : must point at beginning of a valid zstd dictionary.
* return : size of dictionary header (size of magic number + dict ID + entropy tables)
* assumptions : magic number supposed already checked
* and dictSize >= 8 */
size_t ZSTD_loadCEntropy(ZSTD_compressedBlockState_t* bs, void* workspace,
short* offcodeNCount, unsigned* offcodeMaxValue,
const void* const dict, size_t dictSize);
void ZSTD_reset_compressedBlockState(ZSTD_compressedBlockState_t* bs);
/* ==============================================================
* Private declarations
@ -728,6 +1058,7 @@ MEM_STATIC void ZSTD_debugTable(const U32* table, U32 max)
/* ZSTD_getCParamsFromCCtxParams() :
* cParams are built depending on compressionLevel, src size hints,
* LDM and manually set compression parameters.
* Note: srcSizeHint == 0 means 0!
*/
ZSTD_compressionParameters ZSTD_getCParamsFromCCtxParams(
const ZSTD_CCtx_params* CCtxParams, U64 srcSizeHint, size_t dictSize);
@ -740,17 +1071,10 @@ ZSTD_compressionParameters ZSTD_getCParamsFromCCtxParams(
size_t ZSTD_initCStream_internal(ZSTD_CStream* zcs,
const void* dict, size_t dictSize,
const ZSTD_CDict* cdict,
ZSTD_CCtx_params params, unsigned long long pledgedSrcSize);
const ZSTD_CCtx_params* params, unsigned long long pledgedSrcSize);
void ZSTD_resetSeqStore(seqStore_t* ssPtr);
/*! ZSTD_compressStream_generic() :
* Private use only. To be called from zstdmt_compress.c in single-thread mode. */
size_t ZSTD_compressStream_generic(ZSTD_CStream* zcs,
ZSTD_outBuffer* output,
ZSTD_inBuffer* input,
ZSTD_EndDirective const flushMode);
/*! ZSTD_getCParamsFromCDict() :
* as the name implies */
ZSTD_compressionParameters ZSTD_getCParamsFromCDict(const ZSTD_CDict* cdict);
@ -762,7 +1086,7 @@ size_t ZSTD_compressBegin_advanced_internal(ZSTD_CCtx* cctx,
ZSTD_dictContentType_e dictContentType,
ZSTD_dictTableLoadMethod_e dtlm,
const ZSTD_CDict* cdict,
ZSTD_CCtx_params params,
const ZSTD_CCtx_params* params,
unsigned long long pledgedSrcSize);
/* ZSTD_compress_advanced_internal() :
@ -771,13 +1095,13 @@ size_t ZSTD_compress_advanced_internal(ZSTD_CCtx* cctx,
void* dst, size_t dstCapacity,
const void* src, size_t srcSize,
const void* dict,size_t dictSize,
ZSTD_CCtx_params params);
const ZSTD_CCtx_params* params);
/* ZSTD_writeLastEmptyBlock() :
* output an empty Block with end-of-frame mark to complete a frame
* @return : size of data written into `dst` (== ZSTD_blockHeaderSize (defined in zstd_internal.h))
* or an error code if `dstCapcity` is too small (<ZSTD_blockHeaderSize)
* or an error code if `dstCapacity` is too small (<ZSTD_blockHeaderSize)
*/
size_t ZSTD_writeLastEmptyBlock(void* dst, size_t dstCapacity);
@ -794,5 +1118,8 @@ size_t ZSTD_writeLastEmptyBlock(void* dst, size_t dstCapacity);
*/
size_t ZSTD_referenceExternalSequences(ZSTD_CCtx* cctx, rawSeq* seq, size_t nbSeq);
/** ZSTD_cycleLog() :
* condition for correct operation : hashLog > 1 */
U32 ZSTD_cycleLog(U32 hashLog, ZSTD_strategy strat);
#endif /* ZSTD_COMPRESS_H */

158
drivers/filesystems/btrfs/zstd/zstd_compress_literals.c Normal file
View file

@ -0,0 +1,158 @@
/*
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
/*-*************************************
* Dependencies
***************************************/
#include "zstd_compress_literals.h"
size_t ZSTD_noCompressLiterals (void* dst, size_t dstCapacity, const void* src, size_t srcSize)
{
BYTE* const ostart = (BYTE* const)dst;
U32 const flSize = 1 + (srcSize>31) + (srcSize>4095);
RETURN_ERROR_IF(srcSize + flSize > dstCapacity, dstSize_tooSmall, "");
switch(flSize)
{
case 1: /* 2 - 1 - 5 */
ostart[0] = (BYTE)((U32)set_basic + (srcSize<<3));
break;
case 2: /* 2 - 2 - 12 */
MEM_writeLE16(ostart, (U16)((U32)set_basic + (1<<2) + (srcSize<<4)));
break;
case 3: /* 2 - 2 - 20 */
MEM_writeLE32(ostart, (U32)((U32)set_basic + (3<<2) + (srcSize<<4)));
break;
default: /* not necessary : flSize is {1,2,3} */
assert(0);
}
memcpy(ostart + flSize, src, srcSize);
DEBUGLOG(5, "Raw literals: %u -> %u", (U32)srcSize, (U32)(srcSize + flSize));
return srcSize + flSize;
}
size_t ZSTD_compressRleLiteralsBlock (void* dst, size_t dstCapacity, const void* src, size_t srcSize)
{
BYTE* const ostart = (BYTE* const)dst;
U32 const flSize = 1 + (srcSize>31) + (srcSize>4095);
(void)dstCapacity; /* dstCapacity already guaranteed to be >=4, hence large enough */
switch(flSize)
{
case 1: /* 2 - 1 - 5 */
ostart[0] = (BYTE)((U32)set_rle + (srcSize<<3));
break;
case 2: /* 2 - 2 - 12 */
MEM_writeLE16(ostart, (U16)((U32)set_rle + (1<<2) + (srcSize<<4)));
break;
case 3: /* 2 - 2 - 20 */
MEM_writeLE32(ostart, (U32)((U32)set_rle + (3<<2) + (srcSize<<4)));
break;
default: /* not necessary : flSize is {1,2,3} */
assert(0);
}
ostart[flSize] = *(const BYTE*)src;
DEBUGLOG(5, "RLE literals: %u -> %u", (U32)srcSize, (U32)flSize + 1);
return flSize+1;
}
size_t ZSTD_compressLiterals (ZSTD_hufCTables_t const* prevHuf,
ZSTD_hufCTables_t* nextHuf,
ZSTD_strategy strategy, int disableLiteralCompression,
void* dst, size_t dstCapacity,
const void* src, size_t srcSize,
void* entropyWorkspace, size_t entropyWorkspaceSize,
const int bmi2)
{
size_t const minGain = ZSTD_minGain(srcSize, strategy);
size_t const lhSize = 3 + (srcSize >= 1 KB) + (srcSize >= 16 KB);
BYTE* const ostart = (BYTE*)dst;
U32 singleStream = srcSize < 256;
symbolEncodingType_e hType = set_compressed;
size_t cLitSize;
DEBUGLOG(5,"ZSTD_compressLiterals (disableLiteralCompression=%i srcSize=%u)",
disableLiteralCompression, (U32)srcSize);
/* Prepare nextEntropy assuming reusing the existing table */
memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
if (disableLiteralCompression)
return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize);
/* small ? don't even attempt compression (speed opt) */
# define COMPRESS_LITERALS_SIZE_MIN 63
{ size_t const minLitSize = (prevHuf->repeatMode == HUF_repeat_valid) ? 6 : COMPRESS_LITERALS_SIZE_MIN;
if (srcSize <= minLitSize) return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize);
}
RETURN_ERROR_IF(dstCapacity < lhSize+1, dstSize_tooSmall, "not enough space for compression");
{ HUF_repeat repeat = prevHuf->repeatMode;
int const preferRepeat = strategy < ZSTD_lazy ? srcSize <= 1024 : 0;
if (repeat == HUF_repeat_valid && lhSize == 3) singleStream = 1;
cLitSize = singleStream ?
HUF_compress1X_repeat(
ostart+lhSize, dstCapacity-lhSize, src, srcSize,
HUF_SYMBOLVALUE_MAX, HUF_TABLELOG_DEFAULT, entropyWorkspace, entropyWorkspaceSize,
(HUF_CElt*)nextHuf->CTable, &repeat, preferRepeat, bmi2) :
HUF_compress4X_repeat(
ostart+lhSize, dstCapacity-lhSize, src, srcSize,
HUF_SYMBOLVALUE_MAX, HUF_TABLELOG_DEFAULT, entropyWorkspace, entropyWorkspaceSize,
(HUF_CElt*)nextHuf->CTable, &repeat, preferRepeat, bmi2);
if (repeat != HUF_repeat_none) {
/* reused the existing table */
DEBUGLOG(5, "Reusing previous huffman table");
hType = set_repeat;
}
}
if ((cLitSize==0) | (cLitSize >= srcSize - minGain) | ERR_isError(cLitSize)) {
memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
return ZSTD_noCompressLiterals(dst, dstCapacity, src, srcSize);
}
if (cLitSize==1) {
memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
return ZSTD_compressRleLiteralsBlock(dst, dstCapacity, src, srcSize);
}
if (hType == set_compressed) {
/* using a newly constructed table */
nextHuf->repeatMode = HUF_repeat_check;
}
/* Build header */
switch(lhSize)
{
case 3: /* 2 - 2 - 10 - 10 */
{ U32 const lhc = hType + ((!singleStream) << 2) + ((U32)srcSize<<4) + ((U32)cLitSize<<14);
MEM_writeLE24(ostart, lhc);
break;
}
case 4: /* 2 - 2 - 14 - 14 */
{ U32 const lhc = hType + (2 << 2) + ((U32)srcSize<<4) + ((U32)cLitSize<<18);
MEM_writeLE32(ostart, lhc);
break;
}
case 5: /* 2 - 2 - 18 - 18 */
{ U32 const lhc = hType + (3 << 2) + ((U32)srcSize<<4) + ((U32)cLitSize<<22);
MEM_writeLE32(ostart, lhc);
ostart[4] = (BYTE)(cLitSize >> 10);
break;
}
default: /* not possible : lhSize is {3,4,5} */
assert(0);
}
DEBUGLOG(5, "Compressed literals: %u -> %u", (U32)srcSize, (U32)(lhSize+cLitSize));
return lhSize+cLitSize;
}

29
drivers/filesystems/btrfs/zstd/zstd_compress_literals.h Normal file
View file

@ -0,0 +1,29 @@
/*
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
#ifndef ZSTD_COMPRESS_LITERALS_H
#define ZSTD_COMPRESS_LITERALS_H
#include "zstd_compress_internal.h" /* ZSTD_hufCTables_t, ZSTD_minGain() */
size_t ZSTD_noCompressLiterals (void* dst, size_t dstCapacity, const void* src, size_t srcSize);
size_t ZSTD_compressRleLiteralsBlock (void* dst, size_t dstCapacity, const void* src, size_t srcSize);
size_t ZSTD_compressLiterals (ZSTD_hufCTables_t const* prevHuf,
ZSTD_hufCTables_t* nextHuf,
ZSTD_strategy strategy, int disableLiteralCompression,
void* dst, size_t dstCapacity,
const void* src, size_t srcSize,
void* entropyWorkspace, size_t entropyWorkspaceSize,
const int bmi2);
#endif /* ZSTD_COMPRESS_LITERALS_H */

419
drivers/filesystems/btrfs/zstd/zstd_compress_sequences.c Normal file
View file

@ -0,0 +1,419 @@
/*
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
/*-*************************************
* Dependencies
***************************************/
#include "zstd_compress_sequences.h"
/**
* -log2(x / 256) lookup table for x in [0, 256).
* If x == 0: Return 0
* Else: Return floor(-log2(x / 256) * 256)
*/
static unsigned const kInverseProbabilityLog256[256] = {
0, 2048, 1792, 1642, 1536, 1453, 1386, 1329, 1280, 1236, 1197, 1162,
1130, 1100, 1073, 1047, 1024, 1001, 980, 960, 941, 923, 906, 889,
874, 859, 844, 830, 817, 804, 791, 779, 768, 756, 745, 734,
724, 714, 704, 694, 685, 676, 667, 658, 650, 642, 633, 626,
618, 610, 603, 595, 588, 581, 574, 567, 561, 554, 548, 542,
535, 529, 523, 517, 512, 506, 500, 495, 489, 484, 478, 473,
468, 463, 458, 453, 448, 443, 438, 434, 429, 424, 420, 415,
411, 407, 402, 398, 394, 390, 386, 382, 377, 373, 370, 366,
362, 358, 354, 350, 347, 343, 339, 336, 332, 329, 325, 322,
318, 315, 311, 308, 305, 302, 298, 295, 292, 289, 286, 282,
279, 276, 273, 270, 267, 264, 261, 258, 256, 253, 250, 247,
244, 241, 239, 236, 233, 230, 228, 225, 222, 220, 217, 215,
212, 209, 207, 204, 202, 199, 197, 194, 192, 190, 187, 185,
182, 180, 178, 175, 173, 171, 168, 166, 164, 162, 159, 157,
155, 153, 151, 149, 146, 144, 142, 140, 138, 136, 134, 132,
130, 128, 126, 123, 121, 119, 117, 115, 114, 112, 110, 108,
106, 104, 102, 100, 98, 96, 94, 93, 91, 89, 87, 85,
83, 82, 80, 78, 76, 74, 73, 71, 69, 67, 66, 64,
62, 61, 59, 57, 55, 54, 52, 50, 49, 47, 46, 44,
42, 41, 39, 37, 36, 34, 33, 31, 30, 28, 26, 25,
23, 22, 20, 19, 17, 16, 14, 13, 11, 10, 8, 7,
5, 4, 2, 1,
};
static unsigned ZSTD_getFSEMaxSymbolValue(FSE_CTable const* ctable) {
void const* ptr = ctable;
U16 const* u16ptr = (U16 const*)ptr;
U32 const maxSymbolValue = MEM_read16(u16ptr + 1);
return maxSymbolValue;
}
/**
* Returns the cost in bytes of encoding the normalized count header.
* Returns an error if any of the helper functions return an error.
*/
static size_t ZSTD_NCountCost(unsigned const* count, unsigned const max,
size_t const nbSeq, unsigned const FSELog)
{
BYTE wksp[FSE_NCOUNTBOUND];
S16 norm[MaxSeq + 1];
const U32 tableLog = FSE_optimalTableLog(FSELog, nbSeq, max);
FORWARD_IF_ERROR(FSE_normalizeCount(norm, tableLog, count, nbSeq, max), "");
return FSE_writeNCount(wksp, sizeof(wksp), norm, max, tableLog);
}
/**
* Returns the cost in bits of encoding the distribution described by count
* using the entropy bound.
*/
static size_t ZSTD_entropyCost(unsigned const* count, unsigned const max, size_t const total)
{
unsigned cost = 0;
unsigned s;
for (s = 0; s <= max; ++s) {
unsigned norm = (unsigned)((256 * count[s]) / total);
if (count[s] != 0 && norm == 0)
norm = 1;
assert(count[s] < total);
cost += count[s] * kInverseProbabilityLog256[norm];
}
return cost >> 8;
}
/**
* Returns the cost in bits of encoding the distribution in count using ctable.
* Returns an error if ctable cannot represent all the symbols in count.
*/
size_t ZSTD_fseBitCost(
FSE_CTable const* ctable,
unsigned const* count,
unsigned const max)
{
unsigned const kAccuracyLog = 8;
size_t cost = 0;
unsigned s;
FSE_CState_t cstate;
FSE_initCState(&cstate, ctable);
if (ZSTD_getFSEMaxSymbolValue(ctable) < max) {
DEBUGLOG(5, "Repeat FSE_CTable has maxSymbolValue %u < %u",
ZSTD_getFSEMaxSymbolValue(ctable), max);
return ERROR(GENERIC);
}
for (s = 0; s <= max; ++s) {
unsigned const tableLog = cstate.stateLog;
unsigned const badCost = (tableLog + 1) << kAccuracyLog;
unsigned const bitCost = FSE_bitCost(cstate.symbolTT, tableLog, s, kAccuracyLog);
if (count[s] == 0)
continue;
if (bitCost >= badCost) {
DEBUGLOG(5, "Repeat FSE_CTable has Prob[%u] == 0", s);
return ERROR(GENERIC);
}
cost += (size_t)count[s] * bitCost;
}
return cost >> kAccuracyLog;
}
/**
* Returns the cost in bits of encoding the distribution in count using the
* table described by norm. The max symbol support by norm is assumed >= max.
* norm must be valid for every symbol with non-zero probability in count.
*/
size_t ZSTD_crossEntropyCost(short const* norm, unsigned accuracyLog,
unsigned const* count, unsigned const max)
{
unsigned const shift = 8 - accuracyLog;
size_t cost = 0;
unsigned s;
assert(accuracyLog <= 8);
for (s = 0; s <= max; ++s) {
unsigned const normAcc = (norm[s] != -1) ? (unsigned)norm[s] : 1;
unsigned const norm256 = normAcc << shift;
assert(norm256 > 0);
assert(norm256 < 256);
cost += count[s] * kInverseProbabilityLog256[norm256];
}
return cost >> 8;
}
symbolEncodingType_e
ZSTD_selectEncodingType(
FSE_repeat* repeatMode, unsigned const* count, unsigned const max,
size_t const mostFrequent, size_t nbSeq, unsigned const FSELog,
FSE_CTable const* prevCTable,
short const* defaultNorm, U32 defaultNormLog,
ZSTD_defaultPolicy_e const isDefaultAllowed,
ZSTD_strategy const strategy)
{
ZSTD_STATIC_ASSERT(ZSTD_defaultDisallowed == 0 && ZSTD_defaultAllowed != 0);
if (mostFrequent == nbSeq) {
*repeatMode = FSE_repeat_none;
if (isDefaultAllowed && nbSeq <= 2) {
/* Prefer set_basic over set_rle when there are 2 or less symbols,
* since RLE uses 1 byte, but set_basic uses 5-6 bits per symbol.
* If basic encoding isn't possible, always choose RLE.
*/
DEBUGLOG(5, "Selected set_basic");
return set_basic;
}
DEBUGLOG(5, "Selected set_rle");
return set_rle;
}
if (strategy < ZSTD_lazy) {
if (isDefaultAllowed) {
size_t const staticFse_nbSeq_max = 1000;
size_t const mult = 10 - strategy;
size_t const baseLog = 3;
size_t const dynamicFse_nbSeq_min = (((size_t)1 << defaultNormLog) * mult) >> baseLog; /* 28-36 for offset, 56-72 for lengths */
assert(defaultNormLog >= 5 && defaultNormLog <= 6); /* xx_DEFAULTNORMLOG */
assert(mult <= 9 && mult >= 7);
if ( (*repeatMode == FSE_repeat_valid)
&& (nbSeq < staticFse_nbSeq_max) ) {
DEBUGLOG(5, "Selected set_repeat");
return set_repeat;
}
if ( (nbSeq < dynamicFse_nbSeq_min)
|| (mostFrequent < (nbSeq >> (defaultNormLog-1))) ) {
DEBUGLOG(5, "Selected set_basic");
/* The format allows default tables to be repeated, but it isn't useful.
* When using simple heuristics to select encoding type, we don't want
* to confuse these tables with dictionaries. When running more careful
* analysis, we don't need to waste time checking both repeating tables
* and default tables.
*/
*repeatMode = FSE_repeat_none;
return set_basic;
}
}
} else {
size_t const basicCost = isDefaultAllowed ? ZSTD_crossEntropyCost(defaultNorm, defaultNormLog, count, max) : ERROR(GENERIC);
size_t const repeatCost = *repeatMode != FSE_repeat_none ? ZSTD_fseBitCost(prevCTable, count, max) : ERROR(GENERIC);
size_t const NCountCost = ZSTD_NCountCost(count, max, nbSeq, FSELog);
size_t const compressedCost = (NCountCost << 3) + ZSTD_entropyCost(count, max, nbSeq);
if (isDefaultAllowed) {
assert(!ZSTD_isError(basicCost));
assert(!(*repeatMode == FSE_repeat_valid && ZSTD_isError(repeatCost)));
}
assert(!ZSTD_isError(NCountCost));
assert(compressedCost < ERROR(maxCode));
DEBUGLOG(5, "Estimated bit costs: basic=%u\trepeat=%u\tcompressed=%u",
(unsigned)basicCost, (unsigned)repeatCost, (unsigned)compressedCost);
if (basicCost <= repeatCost && basicCost <= compressedCost) {
DEBUGLOG(5, "Selected set_basic");
assert(isDefaultAllowed);
*repeatMode = FSE_repeat_none;
return set_basic;
}
if (repeatCost <= compressedCost) {
DEBUGLOG(5, "Selected set_repeat");
assert(!ZSTD_isError(repeatCost));
return set_repeat;
}
assert(compressedCost < basicCost && compressedCost < repeatCost);
}
DEBUGLOG(5, "Selected set_compressed");
*repeatMode = FSE_repeat_check;
return set_compressed;
}
size_t
ZSTD_buildCTable(void* dst, size_t dstCapacity,
FSE_CTable* nextCTable, U32 FSELog, symbolEncodingType_e type,
unsigned* count, U32 max,
const BYTE* codeTable, size_t nbSeq,
const S16* defaultNorm, U32 defaultNormLog, U32 defaultMax,
const FSE_CTable* prevCTable, size_t prevCTableSize,
void* entropyWorkspace, size_t entropyWorkspaceSize)
{
BYTE* op = (BYTE*)dst;
const BYTE* const oend = op + dstCapacity;
DEBUGLOG(6, "ZSTD_buildCTable (dstCapacity=%u)", (unsigned)dstCapacity);
switch (type) {
case set_rle:
FORWARD_IF_ERROR(FSE_buildCTable_rle(nextCTable, (BYTE)max), "");
RETURN_ERROR_IF(dstCapacity==0, dstSize_tooSmall, "not enough space");
*op = codeTable[0];
return 1;
case set_repeat:
memcpy(nextCTable, prevCTable, prevCTableSize);
return 0;
case set_basic:
FORWARD_IF_ERROR(FSE_buildCTable_wksp(nextCTable, defaultNorm, defaultMax, defaultNormLog, entropyWorkspace, entropyWorkspaceSize), ""); /* note : could be pre-calculated */
return 0;
case set_compressed: {
S16 norm[MaxSeq + 1];
size_t nbSeq_1 = nbSeq;
const U32 tableLog = FSE_optimalTableLog(FSELog, nbSeq, max);
if (count[codeTable[nbSeq-1]] > 1) {
count[codeTable[nbSeq-1]]--;
nbSeq_1--;
}
assert(nbSeq_1 > 1);
FORWARD_IF_ERROR(FSE_normalizeCount(norm, tableLog, count, nbSeq_1, max), "");
{ size_t const NCountSize = FSE_writeNCount(op, oend - op, norm, max, tableLog); /* overflow protected */
FORWARD_IF_ERROR(NCountSize, "FSE_writeNCount failed");
FORWARD_IF_ERROR(FSE_buildCTable_wksp(nextCTable, norm, max, tableLog, entropyWorkspace, entropyWorkspaceSize), "");
return NCountSize;
}
}
default: assert(0); RETURN_ERROR(GENERIC, "impossible to reach");
}
}
FORCE_INLINE_TEMPLATE size_t
ZSTD_encodeSequences_body(
void* dst, size_t dstCapacity,
FSE_CTable const* CTable_MatchLength, BYTE const* mlCodeTable,
FSE_CTable const* CTable_OffsetBits, BYTE const* ofCodeTable,
FSE_CTable const* CTable_LitLength, BYTE const* llCodeTable,
seqDef const* sequences, size_t nbSeq, int longOffsets)
{
BIT_CStream_t blockStream;
FSE_CState_t stateMatchLength;
FSE_CState_t stateOffsetBits;
FSE_CState_t stateLitLength;
RETURN_ERROR_IF(
ERR_isError(BIT_initCStream(&blockStream, dst, dstCapacity)),
dstSize_tooSmall, "not enough space remaining");
DEBUGLOG(6, "available space for bitstream : %i (dstCapacity=%u)",
(int)(blockStream.endPtr - blockStream.startPtr),
(unsigned)dstCapacity);
/* first symbols */
FSE_initCState2(&stateMatchLength, CTable_MatchLength, mlCodeTable[nbSeq-1]);
FSE_initCState2(&stateOffsetBits, CTable_OffsetBits, ofCodeTable[nbSeq-1]);
FSE_initCState2(&stateLitLength, CTable_LitLength, llCodeTable[nbSeq-1]);
BIT_addBits(&blockStream, sequences[nbSeq-1].litLength, LL_bits[llCodeTable[nbSeq-1]]);
if (MEM_32bits()) BIT_flushBits(&blockStream);
BIT_addBits(&blockStream, sequences[nbSeq-1].matchLength, ML_bits[mlCodeTable[nbSeq-1]]);
if (MEM_32bits()) BIT_flushBits(&blockStream);
if (longOffsets) {
U32 const ofBits = ofCodeTable[nbSeq-1];
unsigned const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN-1);
if (extraBits) {
BIT_addBits(&blockStream, sequences[nbSeq-1].offset, extraBits);
BIT_flushBits(&blockStream);
}
BIT_addBits(&blockStream, sequences[nbSeq-1].offset >> extraBits,
ofBits - extraBits);
} else {
BIT_addBits(&blockStream, sequences[nbSeq-1].offset, ofCodeTable[nbSeq-1]);
}
BIT_flushBits(&blockStream);
{ size_t n;
for (n=nbSeq-2 ; n<nbSeq ; n--) { /* intentional underflow */
BYTE const llCode = llCodeTable[n];
BYTE const ofCode = ofCodeTable[n];
BYTE const mlCode = mlCodeTable[n];
U32 const llBits = LL_bits[llCode];
U32 const ofBits = ofCode;
U32 const mlBits = ML_bits[mlCode];
DEBUGLOG(6, "encoding: litlen:%2u - matchlen:%2u - offCode:%7u",
(unsigned)sequences[n].litLength,
(unsigned)sequences[n].matchLength + MINMATCH,
(unsigned)sequences[n].offset);
/* 32b*/ /* 64b*/
/* (7)*/ /* (7)*/
FSE_encodeSymbol(&blockStream, &stateOffsetBits, ofCode); /* 15 */ /* 15 */
FSE_encodeSymbol(&blockStream, &stateMatchLength, mlCode); /* 24 */ /* 24 */
if (MEM_32bits()) BIT_flushBits(&blockStream); /* (7)*/
FSE_encodeSymbol(&blockStream, &stateLitLength, llCode); /* 16 */ /* 33 */
if (MEM_32bits() || (ofBits+mlBits+llBits >= 64-7-(LLFSELog+MLFSELog+OffFSELog)))
BIT_flushBits(&blockStream); /* (7)*/
BIT_addBits(&blockStream, sequences[n].litLength, llBits);
if (MEM_32bits() && ((llBits+mlBits)>24)) BIT_flushBits(&blockStream);
BIT_addBits(&blockStream, sequences[n].matchLength, mlBits);
if (MEM_32bits() || (ofBits+mlBits+llBits > 56)) BIT_flushBits(&blockStream);
if (longOffsets) {
unsigned const extraBits = ofBits - MIN(ofBits, STREAM_ACCUMULATOR_MIN-1);
if (extraBits) {
BIT_addBits(&blockStream, sequences[n].offset, extraBits);
BIT_flushBits(&blockStream); /* (7)*/
}
BIT_addBits(&blockStream, sequences[n].offset >> extraBits,
ofBits - extraBits); /* 31 */
} else {
BIT_addBits(&blockStream, sequences[n].offset, ofBits); /* 31 */
}
BIT_flushBits(&blockStream); /* (7)*/
DEBUGLOG(7, "remaining space : %i", (int)(blockStream.endPtr - blockStream.ptr));
} }
DEBUGLOG(6, "ZSTD_encodeSequences: flushing ML state with %u bits", stateMatchLength.stateLog);
FSE_flushCState(&blockStream, &stateMatchLength);
DEBUGLOG(6, "ZSTD_encodeSequences: flushing Off state with %u bits", stateOffsetBits.stateLog);
FSE_flushCState(&blockStream, &stateOffsetBits);
DEBUGLOG(6, "ZSTD_encodeSequences: flushing LL state with %u bits", stateLitLength.stateLog);
FSE_flushCState(&blockStream, &stateLitLength);
{ size_t const streamSize = BIT_closeCStream(&blockStream);
RETURN_ERROR_IF(streamSize==0, dstSize_tooSmall, "not enough space");
return streamSize;
}
}
static size_t
ZSTD_encodeSequences_default(
void* dst, size_t dstCapacity,
FSE_CTable const* CTable_MatchLength, BYTE const* mlCodeTable,
FSE_CTable const* CTable_OffsetBits, BYTE const* ofCodeTable,
FSE_CTable const* CTable_LitLength, BYTE const* llCodeTable,
seqDef const* sequences, size_t nbSeq, int longOffsets)
{
return ZSTD_encodeSequences_body(dst, dstCapacity,
CTable_MatchLength, mlCodeTable,
CTable_OffsetBits, ofCodeTable,
CTable_LitLength, llCodeTable,
sequences, nbSeq, longOffsets);
}
#if DYNAMIC_BMI2
static TARGET_ATTRIBUTE("bmi2") size_t
ZSTD_encodeSequences_bmi2(
void* dst, size_t dstCapacity,
FSE_CTable const* CTable_MatchLength, BYTE const* mlCodeTable,
FSE_CTable const* CTable_OffsetBits, BYTE const* ofCodeTable,
FSE_CTable const* CTable_LitLength, BYTE const* llCodeTable,
seqDef const* sequences, size_t nbSeq, int longOffsets)
{
return ZSTD_encodeSequences_body(dst, dstCapacity,
CTable_MatchLength, mlCodeTable,
CTable_OffsetBits, ofCodeTable,
CTable_LitLength, llCodeTable,
sequences, nbSeq, longOffsets);
}
#endif
size_t ZSTD_encodeSequences(
void* dst, size_t dstCapacity,
FSE_CTable const* CTable_MatchLength, BYTE const* mlCodeTable,
FSE_CTable const* CTable_OffsetBits, BYTE const* ofCodeTable,
FSE_CTable const* CTable_LitLength, BYTE const* llCodeTable,
seqDef const* sequences, size_t nbSeq, int longOffsets, int bmi2)
{
DEBUGLOG(5, "ZSTD_encodeSequences: dstCapacity = %u", (unsigned)dstCapacity);
#if DYNAMIC_BMI2
if (bmi2) {
return ZSTD_encodeSequences_bmi2(dst, dstCapacity,
CTable_MatchLength, mlCodeTable,
CTable_OffsetBits, ofCodeTable,
CTable_LitLength, llCodeTable,
sequences, nbSeq, longOffsets);
}
#endif
(void)bmi2;
return ZSTD_encodeSequences_default(dst, dstCapacity,
CTable_MatchLength, mlCodeTable,
CTable_OffsetBits, ofCodeTable,
CTable_LitLength, llCodeTable,
sequences, nbSeq, longOffsets);
}

54
drivers/filesystems/btrfs/zstd/zstd_compress_sequences.h Normal file
View file

@ -0,0 +1,54 @@
/*
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
#ifndef ZSTD_COMPRESS_SEQUENCES_H
#define ZSTD_COMPRESS_SEQUENCES_H
#include "fse.h" /* FSE_repeat, FSE_CTable */
#include "zstd_internal.h" /* symbolEncodingType_e, ZSTD_strategy */
typedef enum {
ZSTD_defaultDisallowed = 0,
ZSTD_defaultAllowed = 1
} ZSTD_defaultPolicy_e;
symbolEncodingType_e
ZSTD_selectEncodingType(
FSE_repeat* repeatMode, unsigned const* count, unsigned const max,
size_t const mostFrequent, size_t nbSeq, unsigned const FSELog,
FSE_CTable const* prevCTable,
short const* defaultNorm, U32 defaultNormLog,
ZSTD_defaultPolicy_e const isDefaultAllowed,
ZSTD_strategy const strategy);
size_t
ZSTD_buildCTable(void* dst, size_t dstCapacity,
FSE_CTable* nextCTable, U32 FSELog, symbolEncodingType_e type,
unsigned* count, U32 max,
const BYTE* codeTable, size_t nbSeq,
const S16* defaultNorm, U32 defaultNormLog, U32 defaultMax,
const FSE_CTable* prevCTable, size_t prevCTableSize,
void* entropyWorkspace, size_t entropyWorkspaceSize);
size_t ZSTD_encodeSequences(
void* dst, size_t dstCapacity,
FSE_CTable const* CTable_MatchLength, BYTE const* mlCodeTable,
FSE_CTable const* CTable_OffsetBits, BYTE const* ofCodeTable,
FSE_CTable const* CTable_LitLength, BYTE const* llCodeTable,
seqDef const* sequences, size_t nbSeq, int longOffsets, int bmi2);
size_t ZSTD_fseBitCost(
FSE_CTable const* ctable,
unsigned const* count,
unsigned const max);
size_t ZSTD_crossEntropyCost(short const* norm, unsigned accuracyLog,
unsigned const* count, unsigned const max);
#endif /* ZSTD_COMPRESS_SEQUENCES_H */

845
drivers/filesystems/btrfs/zstd/zstd_compress_superblock.c Normal file
View file

@ -0,0 +1,845 @@
/*
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
/*-*************************************
* Dependencies
***************************************/
#include "zstd_compress_superblock.h"
#include "zstd_internal.h" /* ZSTD_getSequenceLength */
#include "hist.h" /* HIST_countFast_wksp */
#include "zstd_compress_internal.h"
#include "zstd_compress_sequences.h"
#include "zstd_compress_literals.h"
/*-*************************************
* Superblock entropy buffer structs
***************************************/
/** ZSTD_hufCTablesMetadata_t :
* Stores Literals Block Type for a super-block in hType, and
* huffman tree description in hufDesBuffer.
* hufDesSize refers to the size of huffman tree description in bytes.
* This metadata is populated in ZSTD_buildSuperBlockEntropy_literal() */
typedef struct {
symbolEncodingType_e hType;
BYTE hufDesBuffer[500]; /* TODO give name to this value */
size_t hufDesSize;
} ZSTD_hufCTablesMetadata_t;
/** ZSTD_fseCTablesMetadata_t :
* Stores symbol compression modes for a super-block in {ll, ol, ml}Type, and
* fse tables in fseTablesBuffer.
* fseTablesSize refers to the size of fse tables in bytes.
* This metadata is populated in ZSTD_buildSuperBlockEntropy_sequences() */
typedef struct {
symbolEncodingType_e llType;
symbolEncodingType_e ofType;
symbolEncodingType_e mlType;
BYTE fseTablesBuffer[500]; /* TODO give name to this value */
size_t fseTablesSize;
size_t lastCountSize; /* This is to account for bug in 1.3.4. More detail in ZSTD_compressSubBlock_sequences() */
} ZSTD_fseCTablesMetadata_t;
typedef struct {
ZSTD_hufCTablesMetadata_t hufMetadata;
ZSTD_fseCTablesMetadata_t fseMetadata;
} ZSTD_entropyCTablesMetadata_t;
/** ZSTD_buildSuperBlockEntropy_literal() :
* Builds entropy for the super-block literals.
* Stores literals block type (raw, rle, compressed, repeat) and
* huffman description table to hufMetadata.
* @return : size of huffman description table or error code */
static size_t ZSTD_buildSuperBlockEntropy_literal(void* const src, size_t srcSize,
const ZSTD_hufCTables_t* prevHuf,
ZSTD_hufCTables_t* nextHuf,
ZSTD_hufCTablesMetadata_t* hufMetadata,
const int disableLiteralsCompression,
void* workspace, size_t wkspSize)
{
BYTE* const wkspStart = (BYTE*)workspace;
BYTE* const wkspEnd = wkspStart + wkspSize;
BYTE* const countWkspStart = wkspStart;
unsigned* const countWksp = (unsigned*)workspace;
const size_t countWkspSize = (HUF_SYMBOLVALUE_MAX + 1) * sizeof(unsigned);
BYTE* const nodeWksp = countWkspStart + countWkspSize;
const size_t nodeWkspSize = wkspEnd-nodeWksp;
unsigned maxSymbolValue = 255;
unsigned huffLog = HUF_TABLELOG_DEFAULT;
HUF_repeat repeat = prevHuf->repeatMode;
DEBUGLOG(5, "ZSTD_buildSuperBlockEntropy_literal (srcSize=%zu)", srcSize);
/* Prepare nextEntropy assuming reusing the existing table */
memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
if (disableLiteralsCompression) {
DEBUGLOG(5, "set_basic - disabled");
hufMetadata->hType = set_basic;
return 0;
}
/* small ? don't even attempt compression (speed opt) */
# define COMPRESS_LITERALS_SIZE_MIN 63
{ size_t const minLitSize = (prevHuf->repeatMode == HUF_repeat_valid) ? 6 : COMPRESS_LITERALS_SIZE_MIN;
if (srcSize <= minLitSize) {
DEBUGLOG(5, "set_basic - too small");
hufMetadata->hType = set_basic;
return 0;
}
}
/* Scan input and build symbol stats */
{ size_t const largest = HIST_count_wksp (countWksp, &maxSymbolValue, (const BYTE*)src, srcSize, workspace, wkspSize);
FORWARD_IF_ERROR(largest, "HIST_count_wksp failed");
if (largest == srcSize) {
DEBUGLOG(5, "set_rle");
hufMetadata->hType = set_rle;
return 0;
}
if (largest <= (srcSize >> 7)+4) {
DEBUGLOG(5, "set_basic - no gain");
hufMetadata->hType = set_basic;
return 0;
}
}
/* Validate the previous Huffman table */
if (repeat == HUF_repeat_check && !HUF_validateCTable((HUF_CElt const*)prevHuf->CTable, countWksp, maxSymbolValue)) {
repeat = HUF_repeat_none;
}
/* Build Huffman Tree */
memset(nextHuf->CTable, 0, sizeof(nextHuf->CTable));
huffLog = HUF_optimalTableLog(huffLog, srcSize, maxSymbolValue);
{ size_t const maxBits = HUF_buildCTable_wksp((HUF_CElt*)nextHuf->CTable, countWksp,
maxSymbolValue, huffLog,
nodeWksp, nodeWkspSize);
FORWARD_IF_ERROR(maxBits, "HUF_buildCTable_wksp");
huffLog = (U32)maxBits;
{ /* Build and write the CTable */
size_t const newCSize = HUF_estimateCompressedSize(
(HUF_CElt*)nextHuf->CTable, countWksp, maxSymbolValue);
size_t const hSize = HUF_writeCTable(
hufMetadata->hufDesBuffer, sizeof(hufMetadata->hufDesBuffer),
(HUF_CElt*)nextHuf->CTable, maxSymbolValue, huffLog);
/* Check against repeating the previous CTable */
if (repeat != HUF_repeat_none) {
size_t const oldCSize = HUF_estimateCompressedSize(
(HUF_CElt const*)prevHuf->CTable, countWksp, maxSymbolValue);
if (oldCSize < srcSize && (oldCSize <= hSize + newCSize || hSize + 12 >= srcSize)) {
DEBUGLOG(5, "set_repeat - smaller");
memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
hufMetadata->hType = set_repeat;
return 0;
}
}
if (newCSize + hSize >= srcSize) {
DEBUGLOG(5, "set_basic - no gains");
memcpy(nextHuf, prevHuf, sizeof(*prevHuf));
hufMetadata->hType = set_basic;
return 0;
}
DEBUGLOG(5, "set_compressed (hSize=%u)", (U32)hSize);
hufMetadata->hType = set_compressed;
nextHuf->repeatMode = HUF_repeat_check;
return hSize;
}
}
}
/** ZSTD_buildSuperBlockEntropy_sequences() :
* Builds entropy for the super-block sequences.
* Stores symbol compression modes and fse table to fseMetadata.
* @return : size of fse tables or error code */
static size_t ZSTD_buildSuperBlockEntropy_sequences(seqStore_t* seqStorePtr,
const ZSTD_fseCTables_t* prevEntropy,
ZSTD_fseCTables_t* nextEntropy,
const ZSTD_CCtx_params* cctxParams,
ZSTD_fseCTablesMetadata_t* fseMetadata,
void* workspace, size_t wkspSize)
{
BYTE* const wkspStart = (BYTE*)workspace;
BYTE* const wkspEnd = wkspStart + wkspSize;
BYTE* const countWkspStart = wkspStart;
unsigned* const countWksp = (unsigned*)workspace;
const size_t countWkspSize = (MaxSeq + 1) * sizeof(unsigned);
BYTE* const cTableWksp = countWkspStart + countWkspSize;
const size_t cTableWkspSize = wkspEnd-cTableWksp;
ZSTD_strategy const strategy = cctxParams->cParams.strategy;
FSE_CTable* CTable_LitLength = nextEntropy->litlengthCTable;
FSE_CTable* CTable_OffsetBits = nextEntropy->offcodeCTable;
FSE_CTable* CTable_MatchLength = nextEntropy->matchlengthCTable;
const BYTE* const ofCodeTable = seqStorePtr->ofCode;
const BYTE* const llCodeTable = seqStorePtr->llCode;
const BYTE* const mlCodeTable = seqStorePtr->mlCode;
size_t const nbSeq = seqStorePtr->sequences - seqStorePtr->sequencesStart;
BYTE* const ostart = fseMetadata->fseTablesBuffer;
BYTE* const oend = ostart + sizeof(fseMetadata->fseTablesBuffer);
BYTE* op = ostart;
assert(cTableWkspSize >= (1 << MaxFSELog) * sizeof(FSE_FUNCTION_TYPE));
DEBUGLOG(5, "ZSTD_buildSuperBlockEntropy_sequences (nbSeq=%zu)", nbSeq);
memset(workspace, 0, wkspSize);
fseMetadata->lastCountSize = 0;
/* convert length/distances into codes */
ZSTD_seqToCodes(seqStorePtr);
/* build CTable for Literal Lengths */
{ U32 LLtype;
unsigned max = MaxLL;
size_t const mostFrequent = HIST_countFast_wksp(countWksp, &max, llCodeTable, nbSeq, workspace, wkspSize); /* can't fail */
DEBUGLOG(5, "Building LL table");
nextEntropy->litlength_repeatMode = prevEntropy->litlength_repeatMode;
LLtype = ZSTD_selectEncodingType(&nextEntropy->litlength_repeatMode,
countWksp, max, mostFrequent, nbSeq,
LLFSELog, prevEntropy->litlengthCTable,
LL_defaultNorm, LL_defaultNormLog,
ZSTD_defaultAllowed, strategy);
assert(set_basic < set_compressed && set_rle < set_compressed);
assert(!(LLtype < set_compressed && nextEntropy->litlength_repeatMode != FSE_repeat_none)); /* We don't copy tables */
{ size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_LitLength, LLFSELog, (symbolEncodingType_e)LLtype,
countWksp, max, llCodeTable, nbSeq, LL_defaultNorm, LL_defaultNormLog, MaxLL,
prevEntropy->litlengthCTable, sizeof(prevEntropy->litlengthCTable),
cTableWksp, cTableWkspSize);
FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for LitLens failed");
if (LLtype == set_compressed)
fseMetadata->lastCountSize = countSize;
op += countSize;
fseMetadata->llType = (symbolEncodingType_e) LLtype;
} }
/* build CTable for Offsets */
{ U32 Offtype;
unsigned max = MaxOff;
size_t const mostFrequent = HIST_countFast_wksp(countWksp, &max, ofCodeTable, nbSeq, workspace, wkspSize); /* can't fail */
/* We can only use the basic table if max <= DefaultMaxOff, otherwise the offsets are too large */
ZSTD_defaultPolicy_e const defaultPolicy = (max <= DefaultMaxOff) ? ZSTD_defaultAllowed : ZSTD_defaultDisallowed;
DEBUGLOG(5, "Building OF table");
nextEntropy->offcode_repeatMode = prevEntropy->offcode_repeatMode;
Offtype = ZSTD_selectEncodingType(&nextEntropy->offcode_repeatMode,
countWksp, max, mostFrequent, nbSeq,
OffFSELog, prevEntropy->offcodeCTable,
OF_defaultNorm, OF_defaultNormLog,
defaultPolicy, strategy);
assert(!(Offtype < set_compressed && nextEntropy->offcode_repeatMode != FSE_repeat_none)); /* We don't copy tables */
{ size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_OffsetBits, OffFSELog, (symbolEncodingType_e)Offtype,
countWksp, max, ofCodeTable, nbSeq, OF_defaultNorm, OF_defaultNormLog, DefaultMaxOff,
prevEntropy->offcodeCTable, sizeof(prevEntropy->offcodeCTable),
cTableWksp, cTableWkspSize);
FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for Offsets failed");
if (Offtype == set_compressed)
fseMetadata->lastCountSize = countSize;
op += countSize;
fseMetadata->ofType = (symbolEncodingType_e) Offtype;
} }
/* build CTable for MatchLengths */
{ U32 MLtype;
unsigned max = MaxML;
size_t const mostFrequent = HIST_countFast_wksp(countWksp, &max, mlCodeTable, nbSeq, workspace, wkspSize); /* can't fail */
DEBUGLOG(5, "Building ML table (remaining space : %i)", (int)(oend-op));
nextEntropy->matchlength_repeatMode = prevEntropy->matchlength_repeatMode;
MLtype = ZSTD_selectEncodingType(&nextEntropy->matchlength_repeatMode,
countWksp, max, mostFrequent, nbSeq,
MLFSELog, prevEntropy->matchlengthCTable,
ML_defaultNorm, ML_defaultNormLog,
ZSTD_defaultAllowed, strategy);
assert(!(MLtype < set_compressed && nextEntropy->matchlength_repeatMode != FSE_repeat_none)); /* We don't copy tables */
{ size_t const countSize = ZSTD_buildCTable(op, oend - op, CTable_MatchLength, MLFSELog, (symbolEncodingType_e)MLtype,
countWksp, max, mlCodeTable, nbSeq, ML_defaultNorm, ML_defaultNormLog, MaxML,
prevEntropy->matchlengthCTable, sizeof(prevEntropy->matchlengthCTable),
cTableWksp, cTableWkspSize);
FORWARD_IF_ERROR(countSize, "ZSTD_buildCTable for MatchLengths failed");
if (MLtype == set_compressed)
fseMetadata->lastCountSize = countSize;
op += countSize;
fseMetadata->mlType = (symbolEncodingType_e) MLtype;
} }
assert((size_t) (op-ostart) <= sizeof(fseMetadata->fseTablesBuffer));
return op-ostart;
}
/** ZSTD_buildSuperBlockEntropy() :
* Builds entropy for the super-block.
* @return : 0 on success or error code */
static size_t
ZSTD_buildSuperBlockEntropy(seqStore_t* seqStorePtr,
const ZSTD_entropyCTables_t* prevEntropy,
ZSTD_entropyCTables_t* nextEntropy,
const ZSTD_CCtx_params* cctxParams,
ZSTD_entropyCTablesMetadata_t* entropyMetadata,
void* workspace, size_t wkspSize)
{
size_t const litSize = seqStorePtr->lit - seqStorePtr->litStart;
DEBUGLOG(5, "ZSTD_buildSuperBlockEntropy");
entropyMetadata->hufMetadata.hufDesSize =
ZSTD_buildSuperBlockEntropy_literal(seqStorePtr->litStart, litSize,
&prevEntropy->huf, &nextEntropy->huf,
&entropyMetadata->hufMetadata,
ZSTD_disableLiteralsCompression(cctxParams),
workspace, wkspSize);
FORWARD_IF_ERROR(entropyMetadata->hufMetadata.hufDesSize, "ZSTD_buildSuperBlockEntropy_literal failed");
entropyMetadata->fseMetadata.fseTablesSize =
ZSTD_buildSuperBlockEntropy_sequences(seqStorePtr,
&prevEntropy->fse, &nextEntropy->fse,
cctxParams,
&entropyMetadata->fseMetadata,
workspace, wkspSize);
FORWARD_IF_ERROR(entropyMetadata->fseMetadata.fseTablesSize, "ZSTD_buildSuperBlockEntropy_sequences failed");
return 0;
}
/** ZSTD_compressSubBlock_literal() :
* Compresses literals section for a sub-block.
* When we have to write the Huffman table we will sometimes choose a header
* size larger than necessary. This is because we have to pick the header size
* before we know the table size + compressed size, so we have a bound on the
* table size. If we guessed incorrectly, we fall back to uncompressed literals.
*
* We write the header when writeEntropy=1 and set entropyWrriten=1 when we succeeded
* in writing the header, otherwise it is set to 0.
*
* hufMetadata->hType has literals block type info.
* If it is set_basic, all sub-blocks literals section will be Raw_Literals_Block.
* If it is set_rle, all sub-blocks literals section will be RLE_Literals_Block.
* If it is set_compressed, first sub-block's literals section will be Compressed_Literals_Block
* If it is set_compressed, first sub-block's literals section will be Treeless_Literals_Block
* and the following sub-blocks' literals sections will be Treeless_Literals_Block.
* @return : compressed size of literals section of a sub-block
* Or 0 if it unable to compress.
* Or error code */
static size_t ZSTD_compressSubBlock_literal(const HUF_CElt* hufTable,
const ZSTD_hufCTablesMetadata_t* hufMetadata,
const BYTE* literals, size_t litSize,
void* dst, size_t dstSize,
const int bmi2, int writeEntropy, int* entropyWritten)
{
size_t const header = writeEntropy ? 200 : 0;
size_t const lhSize = 3 + (litSize >= (1 KB - header)) + (litSize >= (16 KB - header));
BYTE* const ostart = (BYTE*)dst;
BYTE* const oend = ostart + dstSize;
BYTE* op = ostart + lhSize;
U32 const singleStream = lhSize == 3;
symbolEncodingType_e hType = writeEntropy ? hufMetadata->hType : set_repeat;
size_t cLitSize = 0;
(void)bmi2; /* TODO bmi2... */
DEBUGLOG(5, "ZSTD_compressSubBlock_literal (litSize=%zu, lhSize=%zu, writeEntropy=%d)", litSize, lhSize, writeEntropy);
*entropyWritten = 0;
if (litSize == 0 || hufMetadata->hType == set_basic) {
DEBUGLOG(5, "ZSTD_compressSubBlock_literal using raw literal");
return ZSTD_noCompressLiterals(dst, dstSize, literals, litSize);
} else if (hufMetadata->hType == set_rle) {
DEBUGLOG(5, "ZSTD_compressSubBlock_literal using rle literal");
return ZSTD_compressRleLiteralsBlock(dst, dstSize, literals, litSize);
}
assert(litSize > 0);
assert(hufMetadata->hType == set_compressed || hufMetadata->hType == set_repeat);
if (writeEntropy && hufMetadata->hType == set_compressed) {
memcpy(op, hufMetadata->hufDesBuffer, hufMetadata->hufDesSize);
op += hufMetadata->hufDesSize;
cLitSize += hufMetadata->hufDesSize;
DEBUGLOG(5, "ZSTD_compressSubBlock_literal (hSize=%zu)", hufMetadata->hufDesSize);
}
/* TODO bmi2 */
{ const size_t cSize = singleStream ? HUF_compress1X_usingCTable(op, oend-op, literals, litSize, hufTable)
: HUF_compress4X_usingCTable(op, oend-op, literals, litSize, hufTable);
op += cSize;
cLitSize += cSize;
if (cSize == 0 || ERR_isError(cSize)) {
DEBUGLOG(5, "Failed to write entropy tables %s", ZSTD_getErrorName(cSize));
return 0;
}
/* If we expand and we aren't writing a header then emit uncompressed */
if (!writeEntropy && cLitSize >= litSize) {
DEBUGLOG(5, "ZSTD_compressSubBlock_literal using raw literal because uncompressible");
return ZSTD_noCompressLiterals(dst, dstSize, literals, litSize);
}
/* If we are writing headers then allow expansion that doesn't change our header size. */
if (lhSize < (size_t)(3 + (cLitSize >= 1 KB) + (cLitSize >= 16 KB))) {
assert(cLitSize > litSize);
DEBUGLOG(5, "Literals expanded beyond allowed header size");
return ZSTD_noCompressLiterals(dst, dstSize, literals, litSize);
}
DEBUGLOG(5, "ZSTD_compressSubBlock_literal (cSize=%zu)", cSize);
}
/* Build header */
switch(lhSize)
{
case 3: /* 2 - 2 - 10 - 10 */
{ U32 const lhc = hType + ((!singleStream) << 2) + ((U32)litSize<<4) + ((U32)cLitSize<<14);
MEM_writeLE24(ostart, lhc);
break;
}
case 4: /* 2 - 2 - 14 - 14 */
{ U32 const lhc = hType + (2 << 2) + ((U32)litSize<<4) + ((U32)cLitSize<<18);
MEM_writeLE32(ostart, lhc);
break;
}
case 5: /* 2 - 2 - 18 - 18 */
{ U32 const lhc = hType + (3 << 2) + ((U32)litSize<<4) + ((U32)cLitSize<<22);
MEM_writeLE32(ostart, lhc);
ostart[4] = (BYTE)(cLitSize >> 10);
break;
}
default: /* not possible : lhSize is {3,4,5} */
assert(0);
}
*entropyWritten = 1;
DEBUGLOG(5, "Compressed literals: %u -> %u", (U32)litSize, (U32)(op-ostart));
return op-ostart;
}
static size_t ZSTD_seqDecompressedSize(seqStore_t const* seqStore, const seqDef* sequences, size_t nbSeq, size_t litSize, int lastSequence) {
const seqDef* const sstart = sequences;
const seqDef* const send = sequences + nbSeq;
const seqDef* sp = sstart;
size_t matchLengthSum = 0;
size_t litLengthSum = 0;
while (send-sp > 0) {
ZSTD_sequenceLength const seqLen = ZSTD_getSequenceLength(seqStore, sp);
litLengthSum += seqLen.litLength;
matchLengthSum += seqLen.matchLength;
sp++;
}
assert(litLengthSum <= litSize);
if (!lastSequence) {
assert(litLengthSum == litSize);
}
return matchLengthSum + litSize;
}
/** ZSTD_compressSubBlock_sequences() :
* Compresses sequences section for a sub-block.
* fseMetadata->llType, fseMetadata->ofType, and fseMetadata->mlType have
* symbol compression modes for the super-block.
* The first successfully compressed block will have these in its header.
* We set entropyWritten=1 when we succeed in compressing the sequences.
* The following sub-blocks will always have repeat mode.
* @return : compressed size of sequences section of a sub-block
* Or 0 if it is unable to compress
* Or error code. */
static size_t ZSTD_compressSubBlock_sequences(const ZSTD_fseCTables_t* fseTables,
const ZSTD_fseCTablesMetadata_t* fseMetadata,
const seqDef* sequences, size_t nbSeq,
const BYTE* llCode, const BYTE* mlCode, const BYTE* ofCode,
const ZSTD_CCtx_params* cctxParams,
void* dst, size_t dstCapacity,
const int bmi2, int writeEntropy, int* entropyWritten)
{
const int longOffsets = cctxParams->cParams.windowLog > STREAM_ACCUMULATOR_MIN;
BYTE* const ostart = (BYTE*)dst;
BYTE* const oend = ostart + dstCapacity;
BYTE* op = ostart;
BYTE* seqHead;
DEBUGLOG(5, "ZSTD_compressSubBlock_sequences (nbSeq=%zu, writeEntropy=%d, longOffsets=%d)", nbSeq, writeEntropy, longOffsets);
*entropyWritten = 0;
/* Sequences Header */
RETURN_ERROR_IF((oend-op) < 3 /*max nbSeq Size*/ + 1 /*seqHead*/,
dstSize_tooSmall, "");
if (nbSeq < 0x7F)
*op++ = (BYTE)nbSeq;
else if (nbSeq < LONGNBSEQ)
op[0] = (BYTE)((nbSeq>>8) + 0x80), op[1] = (BYTE)nbSeq, op+=2;
else
op[0]=0xFF, MEM_writeLE16(op+1, (U16)(nbSeq - LONGNBSEQ)), op+=3;
if (nbSeq==0) {
return op - ostart;
}
/* seqHead : flags for FSE encoding type */
seqHead = op++;
DEBUGLOG(5, "ZSTD_compressSubBlock_sequences (seqHeadSize=%u)", (unsigned)(op-ostart));
if (writeEntropy) {
const U32 LLtype = fseMetadata->llType;
const U32 Offtype = fseMetadata->ofType;
const U32 MLtype = fseMetadata->mlType;
DEBUGLOG(5, "ZSTD_compressSubBlock_sequences (fseTablesSize=%zu)", fseMetadata->fseTablesSize);
*seqHead = (BYTE)((LLtype<<6) + (Offtype<<4) + (MLtype<<2));
memcpy(op, fseMetadata->fseTablesBuffer, fseMetadata->fseTablesSize);
op += fseMetadata->fseTablesSize;
} else {
const U32 repeat = set_repeat;
*seqHead = (BYTE)((repeat<<6) + (repeat<<4) + (repeat<<2));
}
{ size_t const bitstreamSize = ZSTD_encodeSequences(
op, oend - op,
fseTables->matchlengthCTable, mlCode,
fseTables->offcodeCTable, ofCode,
fseTables->litlengthCTable, llCode,
sequences, nbSeq,
longOffsets, bmi2);
FORWARD_IF_ERROR(bitstreamSize, "ZSTD_encodeSequences failed");
op += bitstreamSize;
/* zstd versions <= 1.3.4 mistakenly report corruption when
* FSE_readNCount() receives a buffer < 4 bytes.
* Fixed by https://github.com/facebook/zstd/pull/1146.
* This can happen when the last set_compressed table present is 2
* bytes and the bitstream is only one byte.
* In this exceedingly rare case, we will simply emit an uncompressed
* block, since it isn't worth optimizing.
*/
#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
if (writeEntropy && fseMetadata->lastCountSize && fseMetadata->lastCountSize + bitstreamSize < 4) {
/* NCountSize >= 2 && bitstreamSize > 0 ==> lastCountSize == 3 */
assert(fseMetadata->lastCountSize + bitstreamSize == 3);
DEBUGLOG(5, "Avoiding bug in zstd decoder in versions <= 1.3.4 by "
"emitting an uncompressed block.");
return 0;
}
#endif
DEBUGLOG(5, "ZSTD_compressSubBlock_sequences (bitstreamSize=%zu)", bitstreamSize);
}
/* zstd versions <= 1.4.0 mistakenly report error when
* sequences section body size is less than 3 bytes.
* Fixed by https://github.com/facebook/zstd/pull/1664.
* This can happen when the previous sequences section block is compressed
* with rle mode and the current block's sequences section is compressed
* with repeat mode where sequences section body size can be 1 byte.
*/
#ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
if (op-seqHead < 4) {
DEBUGLOG(5, "Avoiding bug in zstd decoder in versions <= 1.4.0 by emitting "
"an uncompressed block when sequences are < 4 bytes");
return 0;
}
#endif
*entropyWritten = 1;
return op - ostart;
}
/** ZSTD_compressSubBlock() :
* Compresses a single sub-block.
* @return : compressed size of the sub-block
* Or 0 if it failed to compress. */
static size_t ZSTD_compressSubBlock(const ZSTD_entropyCTables_t* entropy,
const ZSTD_entropyCTablesMetadata_t* entropyMetadata,
const seqDef* sequences, size_t nbSeq,
const BYTE* literals, size_t litSize,
const BYTE* llCode, const BYTE* mlCode, const BYTE* ofCode,
const ZSTD_CCtx_params* cctxParams,
void* dst, size_t dstCapacity,
const int bmi2,
int writeLitEntropy, int writeSeqEntropy,
int* litEntropyWritten, int* seqEntropyWritten,
U32 lastBlock)
{
BYTE* const ostart = (BYTE*)dst;
BYTE* const oend = ostart + dstCapacity;
BYTE* op = ostart + ZSTD_blockHeaderSize;
DEBUGLOG(5, "ZSTD_compressSubBlock (litSize=%zu, nbSeq=%zu, writeLitEntropy=%d, writeSeqEntropy=%d, lastBlock=%d)",
litSize, nbSeq, writeLitEntropy, writeSeqEntropy, lastBlock);
{ size_t cLitSize = ZSTD_compressSubBlock_literal((const HUF_CElt*)entropy->huf.CTable,
&entropyMetadata->hufMetadata, literals, litSize,
op, oend-op, bmi2, writeLitEntropy, litEntropyWritten);
FORWARD_IF_ERROR(cLitSize, "ZSTD_compressSubBlock_literal failed");
if (cLitSize == 0) return 0;
op += cLitSize;
}
{ size_t cSeqSize = ZSTD_compressSubBlock_sequences(&entropy->fse,
&entropyMetadata->fseMetadata,
sequences, nbSeq,
llCode, mlCode, ofCode,
cctxParams,
op, oend-op,
bmi2, writeSeqEntropy, seqEntropyWritten);
FORWARD_IF_ERROR(cSeqSize, "ZSTD_compressSubBlock_sequences failed");
if (cSeqSize == 0) return 0;
op += cSeqSize;
}
/* Write block header */
{ size_t cSize = (op-ostart)-ZSTD_blockHeaderSize;
U32 const cBlockHeader24 = lastBlock + (((U32)bt_compressed)<<1) + (U32)(cSize << 3);
MEM_writeLE24(ostart, cBlockHeader24);
}
return op-ostart;
}
static size_t ZSTD_estimateSubBlockSize_literal(const BYTE* literals, size_t litSize,
const ZSTD_hufCTables_t* huf,
const ZSTD_hufCTablesMetadata_t* hufMetadata,
void* workspace, size_t wkspSize,
int writeEntropy)
{
unsigned* const countWksp = (unsigned*)workspace;
unsigned maxSymbolValue = 255;
size_t literalSectionHeaderSize = 3; /* Use hard coded size of 3 bytes */
if (hufMetadata->hType == set_basic) return litSize;
else if (hufMetadata->hType == set_rle) return 1;
else if (hufMetadata->hType == set_compressed || hufMetadata->hType == set_repeat) {
size_t const largest = HIST_count_wksp (countWksp, &maxSymbolValue, (const BYTE*)literals, litSize, workspace, wkspSize);
if (ZSTD_isError(largest)) return litSize;
{ size_t cLitSizeEstimate = HUF_estimateCompressedSize((const HUF_CElt*)huf->CTable, countWksp, maxSymbolValue);
if (writeEntropy) cLitSizeEstimate += hufMetadata->hufDesSize;
return cLitSizeEstimate + literalSectionHeaderSize;
} }
assert(0); /* impossible */
return 0;
}
static size_t ZSTD_estimateSubBlockSize_symbolType(symbolEncodingType_e type,
const BYTE* codeTable, unsigned maxCode,
size_t nbSeq, const FSE_CTable* fseCTable,
const U32* additionalBits,
short const* defaultNorm, U32 defaultNormLog,
void* workspace, size_t wkspSize)
{
unsigned* const countWksp = (unsigned*)workspace;
const BYTE* ctp = codeTable;
const BYTE* const ctStart = ctp;
const BYTE* const ctEnd = ctStart + nbSeq;
size_t cSymbolTypeSizeEstimateInBits = 0;
unsigned max = maxCode;
HIST_countFast_wksp(countWksp, &max, codeTable, nbSeq, workspace, wkspSize); /* can't fail */
if (type == set_basic) {
cSymbolTypeSizeEstimateInBits = ZSTD_crossEntropyCost(defaultNorm, defaultNormLog, countWksp, max);
} else if (type == set_rle) {
cSymbolTypeSizeEstimateInBits = 0;
} else if (type == set_compressed || type == set_repeat) {
cSymbolTypeSizeEstimateInBits = ZSTD_fseBitCost(fseCTable, countWksp, max);
}
if (ZSTD_isError(cSymbolTypeSizeEstimateInBits)) return nbSeq * 10;
while (ctp < ctEnd) {
if (additionalBits) cSymbolTypeSizeEstimateInBits += additionalBits[*ctp];
else cSymbolTypeSizeEstimateInBits += *ctp; /* for offset, offset code is also the number of additional bits */
ctp++;
}
return cSymbolTypeSizeEstimateInBits / 8;
}
static size_t ZSTD_estimateSubBlockSize_sequences(const BYTE* ofCodeTable,
const BYTE* llCodeTable,
const BYTE* mlCodeTable,
size_t nbSeq,
const ZSTD_fseCTables_t* fseTables,
const ZSTD_fseCTablesMetadata_t* fseMetadata,
void* workspace, size_t wkspSize,
int writeEntropy)
{
size_t sequencesSectionHeaderSize = 3; /* Use hard coded size of 3 bytes */
size_t cSeqSizeEstimate = 0;
cSeqSizeEstimate += ZSTD_estimateSubBlockSize_symbolType(fseMetadata->ofType, ofCodeTable, MaxOff,
nbSeq, fseTables->offcodeCTable, NULL,
OF_defaultNorm, OF_defaultNormLog,
workspace, wkspSize);
cSeqSizeEstimate += ZSTD_estimateSubBlockSize_symbolType(fseMetadata->llType, llCodeTable, MaxLL,
nbSeq, fseTables->litlengthCTable, LL_bits,
LL_defaultNorm, LL_defaultNormLog,
workspace, wkspSize);
cSeqSizeEstimate += ZSTD_estimateSubBlockSize_symbolType(fseMetadata->mlType, mlCodeTable, MaxML,
nbSeq, fseTables->matchlengthCTable, ML_bits,
ML_defaultNorm, ML_defaultNormLog,
workspace, wkspSize);
if (writeEntropy) cSeqSizeEstimate += fseMetadata->fseTablesSize;
return cSeqSizeEstimate + sequencesSectionHeaderSize;
}
static size_t ZSTD_estimateSubBlockSize(const BYTE* literals, size_t litSize,
const BYTE* ofCodeTable,
const BYTE* llCodeTable,
const BYTE* mlCodeTable,
size_t nbSeq,
const ZSTD_entropyCTables_t* entropy,
const ZSTD_entropyCTablesMetadata_t* entropyMetadata,
void* workspace, size_t wkspSize,
int writeLitEntropy, int writeSeqEntropy) {
size_t cSizeEstimate = 0;
cSizeEstimate += ZSTD_estimateSubBlockSize_literal(literals, litSize,
&entropy->huf, &entropyMetadata->hufMetadata,
workspace, wkspSize, writeLitEntropy);
cSizeEstimate += ZSTD_estimateSubBlockSize_sequences(ofCodeTable, llCodeTable, mlCodeTable,
nbSeq, &entropy->fse, &entropyMetadata->fseMetadata,
workspace, wkspSize, writeSeqEntropy);
return cSizeEstimate + ZSTD_blockHeaderSize;
}
static int ZSTD_needSequenceEntropyTables(ZSTD_fseCTablesMetadata_t const* fseMetadata)
{
if (fseMetadata->llType == set_compressed || fseMetadata->llType == set_rle)
return 1;
if (fseMetadata->mlType == set_compressed || fseMetadata->mlType == set_rle)
return 1;
if (fseMetadata->ofType == set_compressed || fseMetadata->ofType == set_rle)
return 1;
return 0;
}
/** ZSTD_compressSubBlock_multi() :
* Breaks super-block into multiple sub-blocks and compresses them.
* Entropy will be written to the first block.
* The following blocks will use repeat mode to compress.
* All sub-blocks are compressed blocks (no raw or rle blocks).
* @return : compressed size of the super block (which is multiple ZSTD blocks)
* Or 0 if it failed to compress. */
static size_t ZSTD_compressSubBlock_multi(const seqStore_t* seqStorePtr,
const ZSTD_compressedBlockState_t* prevCBlock,
ZSTD_compressedBlockState_t* nextCBlock,
const ZSTD_entropyCTablesMetadata_t* entropyMetadata,
const ZSTD_CCtx_params* cctxParams,
void* dst, size_t dstCapacity,
const void* src, size_t srcSize,
const int bmi2, U32 lastBlock,
void* workspace, size_t wkspSize)
{
const seqDef* const sstart = seqStorePtr->sequencesStart;
const seqDef* const send = seqStorePtr->sequences;
const seqDef* sp = sstart;
const BYTE* const lstart = seqStorePtr->litStart;
const BYTE* const lend = seqStorePtr->lit;
const BYTE* lp = lstart;
BYTE const* ip = (BYTE const*)src;
BYTE const* const iend = ip + srcSize;
BYTE* const ostart = (BYTE*)dst;
BYTE* const oend = ostart + dstCapacity;
BYTE* op = ostart;
const BYTE* llCodePtr = seqStorePtr->llCode;
const BYTE* mlCodePtr = seqStorePtr->mlCode;
const BYTE* ofCodePtr = seqStorePtr->ofCode;
size_t targetCBlockSize = cctxParams->targetCBlockSize;
size_t litSize, seqCount;
int writeLitEntropy = entropyMetadata->hufMetadata.hType == set_compressed;
int writeSeqEntropy = 1;
int lastSequence = 0;
DEBUGLOG(5, "ZSTD_compressSubBlock_multi (litSize=%u, nbSeq=%u)",
(unsigned)(lend-lp), (unsigned)(send-sstart));
litSize = 0;
seqCount = 0;
do {
size_t cBlockSizeEstimate = 0;
if (sstart == send) {
lastSequence = 1;
} else {
const seqDef* const sequence = sp + seqCount;
lastSequence = sequence == send - 1;
litSize += ZSTD_getSequenceLength(seqStorePtr, sequence).litLength;
seqCount++;
}
if (lastSequence) {
assert(lp <= lend);
assert(litSize <= (size_t)(lend - lp));
litSize = (size_t)(lend - lp);
}
/* I think there is an optimization opportunity here.
* Calling ZSTD_estimateSubBlockSize for every sequence can be wasteful
* since it recalculates estimate from scratch.
* For example, it would recount literal distribution and symbol codes everytime.
*/
cBlockSizeEstimate = ZSTD_estimateSubBlockSize(lp, litSize, ofCodePtr, llCodePtr, mlCodePtr, seqCount,
&nextCBlock->entropy, entropyMetadata,
workspace, wkspSize, writeLitEntropy, writeSeqEntropy);
if (cBlockSizeEstimate > targetCBlockSize || lastSequence) {
int litEntropyWritten = 0;
int seqEntropyWritten = 0;
const size_t decompressedSize = ZSTD_seqDecompressedSize(seqStorePtr, sp, seqCount, litSize, lastSequence);
const size_t cSize = ZSTD_compressSubBlock(&nextCBlock->entropy, entropyMetadata,
sp, seqCount,
lp, litSize,
llCodePtr, mlCodePtr, ofCodePtr,
cctxParams,
op, oend-op,
bmi2, writeLitEntropy, writeSeqEntropy,
&litEntropyWritten, &seqEntropyWritten,
lastBlock && lastSequence);
FORWARD_IF_ERROR(cSize, "ZSTD_compressSubBlock failed");
if (cSize > 0 && cSize < decompressedSize) {
DEBUGLOG(5, "Committed the sub-block");
assert(ip + decompressedSize <= iend);
ip += decompressedSize;
sp += seqCount;
lp += litSize;
op += cSize;
llCodePtr += seqCount;
mlCodePtr += seqCount;
ofCodePtr += seqCount;
litSize = 0;
seqCount = 0;
/* Entropy only needs to be written once */
if (litEntropyWritten) {
writeLitEntropy = 0;
}
if (seqEntropyWritten) {
writeSeqEntropy = 0;
}
}
}
} while (!lastSequence);
if (writeLitEntropy) {
DEBUGLOG(5, "ZSTD_compressSubBlock_multi has literal entropy tables unwritten");
memcpy(&nextCBlock->entropy.huf, &prevCBlock->entropy.huf, sizeof(prevCBlock->entropy.huf));
}
if (writeSeqEntropy && ZSTD_needSequenceEntropyTables(&entropyMetadata->fseMetadata)) {
/* If we haven't written our entropy tables, then we've violated our contract and
* must emit an uncompressed block.
*/
DEBUGLOG(5, "ZSTD_compressSubBlock_multi has sequence entropy tables unwritten");
return 0;
}
if (ip < iend) {
size_t const cSize = ZSTD_noCompressBlock(op, oend - op, ip, iend - ip, lastBlock);
DEBUGLOG(5, "ZSTD_compressSubBlock_multi last sub-block uncompressed, %zu bytes", (size_t)(iend - ip));
FORWARD_IF_ERROR(cSize, "ZSTD_noCompressBlock failed");
assert(cSize != 0);
op += cSize;
/* We have to regenerate the repcodes because we've skipped some sequences */
if (sp < send) {
seqDef const* seq;
repcodes_t rep;
memcpy(&rep, prevCBlock->rep, sizeof(rep));
for (seq = sstart; seq < sp; ++seq) {
rep = ZSTD_updateRep(rep.rep, seq->offset - 1, ZSTD_getSequenceLength(seqStorePtr, seq).litLength == 0);
}
memcpy(nextCBlock->rep, &rep, sizeof(rep));
}
}
DEBUGLOG(5, "ZSTD_compressSubBlock_multi compressed");
return op-ostart;
}
size_t ZSTD_compressSuperBlock(ZSTD_CCtx* zc,
void* dst, size_t dstCapacity,
void const* src, size_t srcSize,
unsigned lastBlock) {
ZSTD_entropyCTablesMetadata_t entropyMetadata;
FORWARD_IF_ERROR(ZSTD_buildSuperBlockEntropy(&zc->seqStore,
&zc->blockState.prevCBlock->entropy,
&zc->blockState.nextCBlock->entropy,
&zc->appliedParams,
&entropyMetadata,
zc->entropyWorkspace, HUF_WORKSPACE_SIZE /* statically allocated in resetCCtx */), "");
return ZSTD_compressSubBlock_multi(&zc->seqStore,
zc->blockState.prevCBlock,
zc->blockState.nextCBlock,
&entropyMetadata,
&zc->appliedParams,
dst, dstCapacity,
src, srcSize,
zc->bmi2, lastBlock,
zc->entropyWorkspace, HUF_WORKSPACE_SIZE /* statically allocated in resetCCtx */);
}

32
drivers/filesystems/btrfs/zstd/zstd_compress_superblock.h Normal file
View file

@ -0,0 +1,32 @@
/*
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
#ifndef ZSTD_COMPRESS_ADVANCED_H
#define ZSTD_COMPRESS_ADVANCED_H
/*-*************************************
* Dependencies
***************************************/
#include "zstd.h" /* ZSTD_CCtx */
/*-*************************************
* Target Compressed Block Size
***************************************/
/* ZSTD_compressSuperBlock() :
* Used to compress a super block when targetCBlockSize is being used.
* The given block will be compressed into multiple sub blocks that are around targetCBlockSize. */
size_t ZSTD_compressSuperBlock(ZSTD_CCtx* zc,
void* dst, size_t dstCapacity,
void const* src, size_t srcSize,
unsigned lastBlock);
#endif /* ZSTD_COMPRESS_ADVANCED_H */

525
drivers/filesystems/btrfs/zstd/zstd_cwksp.h Normal file
View file

@ -0,0 +1,525 @@
/*
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
#ifndef ZSTD_CWKSP_H
#define ZSTD_CWKSP_H
/*-*************************************
* Dependencies
***************************************/
#include "zstd_internal.h"
#if defined (__cplusplus)
extern "C" {
#endif
/*-*************************************
* Constants
***************************************/
/* Since the workspace is effectively its own little malloc implementation /
* arena, when we run under ASAN, we should similarly insert redzones between
* each internal element of the workspace, so ASAN will catch overruns that
* reach outside an object but that stay inside the workspace.
*
* This defines the size of that redzone.
*/
#ifndef ZSTD_CWKSP_ASAN_REDZONE_SIZE
#define ZSTD_CWKSP_ASAN_REDZONE_SIZE 128
#endif
/*-*************************************
* Structures
***************************************/
typedef enum {
ZSTD_cwksp_alloc_objects,
ZSTD_cwksp_alloc_buffers,
ZSTD_cwksp_alloc_aligned
} ZSTD_cwksp_alloc_phase_e;
/**
* Zstd fits all its internal datastructures into a single continuous buffer,
* so that it only needs to perform a single OS allocation (or so that a buffer
* can be provided to it and it can perform no allocations at all). This buffer
* is called the workspace.
*
* Several optimizations complicate that process of allocating memory ranges
* from this workspace for each internal datastructure:
*
* - These different internal datastructures have different setup requirements:
*
* - The static objects need to be cleared once and can then be trivially
* reused for each compression.
*
* - Various buffers don't need to be initialized at all--they are always
* written into before they're read.
*
* - The matchstate tables have a unique requirement that they don't need
* their memory to be totally cleared, but they do need the memory to have
* some bound, i.e., a guarantee that all values in the memory they've been
* allocated is less than some maximum value (which is the starting value
* for the indices that they will then use for compression). When this
* guarantee is provided to them, they can use the memory without any setup
* work. When it can't, they have to clear the area.
*
* - These buffers also have different alignment requirements.
*
* - We would like to reuse the objects in the workspace for multiple
* compressions without having to perform any expensive reallocation or
* reinitialization work.
*
* - We would like to be able to efficiently reuse the workspace across
* multiple compressions **even when the compression parameters change** and
* we need to resize some of the objects (where possible).
*
* To attempt to manage this buffer, given these constraints, the ZSTD_cwksp
* abstraction was created. It works as follows:
*
* Workspace Layout:
*
* [ ... workspace ... ]
* [objects][tables ... ->] free space [<- ... aligned][<- ... buffers]
*
* The various objects that live in the workspace are divided into the
* following categories, and are allocated separately:
*
* - Static objects: this is optionally the enclosing ZSTD_CCtx or ZSTD_CDict,
* so that literally everything fits in a single buffer. Note: if present,
* this must be the first object in the workspace, since ZSTD_free{CCtx,
* CDict}() rely on a pointer comparison to see whether one or two frees are
* required.
*
* - Fixed size objects: these are fixed-size, fixed-count objects that are
* nonetheless "dynamically" allocated in the workspace so that we can
* control how they're initialized separately from the broader ZSTD_CCtx.
* Examples:
* - Entropy Workspace
* - 2 x ZSTD_compressedBlockState_t
* - CDict dictionary contents
*
* - Tables: these are any of several different datastructures (hash tables,
* chain tables, binary trees) that all respect a common format: they are
* uint32_t arrays, all of whose values are between 0 and (nextSrc - base).
* Their sizes depend on the cparams.
*
* - Aligned: these buffers are used for various purposes that require 4 byte
* alignment, but don't require any initialization before they're used.
*
* - Buffers: these buffers are used for various purposes that don't require
* any alignment or initialization before they're used. This means they can
* be moved around at no cost for a new compression.
*
* Allocating Memory:
*
* The various types of objects must be allocated in order, so they can be
* correctly packed into the workspace buffer. That order is:
*
* 1. Objects
* 2. Buffers
* 3. Aligned
* 4. Tables
*
* Attempts to reserve objects of different types out of order will fail.
*/
typedef struct {
void* workspace;
void* workspaceEnd;
void* objectEnd;
void* tableEnd;
void* tableValidEnd;
void* allocStart;
int allocFailed;
int workspaceOversizedDuration;
ZSTD_cwksp_alloc_phase_e phase;
} ZSTD_cwksp;
/*-*************************************
* Functions
***************************************/
MEM_STATIC size_t ZSTD_cwksp_available_space(ZSTD_cwksp* ws);
MEM_STATIC void ZSTD_cwksp_assert_internal_consistency(ZSTD_cwksp* ws) {
(void)ws;
assert(ws->workspace <= ws->objectEnd);
assert(ws->objectEnd <= ws->tableEnd);
assert(ws->objectEnd <= ws->tableValidEnd);
assert(ws->tableEnd <= ws->allocStart);
assert(ws->tableValidEnd <= ws->allocStart);
assert(ws->allocStart <= ws->workspaceEnd);
}
/**
* Align must be a power of 2.
*/
MEM_STATIC size_t ZSTD_cwksp_align(size_t size, size_t const align) {
size_t const mask = align - 1;
assert((align & mask) == 0);
return (size + mask) & ~mask;
}
/**
* Use this to determine how much space in the workspace we will consume to
* allocate this object. (Normally it should be exactly the size of the object,
* but under special conditions, like ASAN, where we pad each object, it might
* be larger.)
*
* Since tables aren't currently redzoned, you don't need to call through this
* to figure out how much space you need for the matchState tables. Everything
* else is though.
*/
MEM_STATIC size_t ZSTD_cwksp_alloc_size(size_t size) {
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
return size + 2 * ZSTD_CWKSP_ASAN_REDZONE_SIZE;
#else
return size;
#endif
}
MEM_STATIC void ZSTD_cwksp_internal_advance_phase(
ZSTD_cwksp* ws, ZSTD_cwksp_alloc_phase_e phase) {
assert(phase >= ws->phase);
if (phase > ws->phase) {
if (ws->phase < ZSTD_cwksp_alloc_buffers &&
phase >= ZSTD_cwksp_alloc_buffers) {
ws->tableValidEnd = ws->objectEnd;
}
if (ws->phase < ZSTD_cwksp_alloc_aligned &&
phase >= ZSTD_cwksp_alloc_aligned) {
/* If unaligned allocations down from a too-large top have left us
* unaligned, we need to realign our alloc ptr. Technically, this
* can consume space that is unaccounted for in the neededSpace
* calculation. However, I believe this can only happen when the
* workspace is too large, and specifically when it is too large
* by a larger margin than the space that will be consumed. */
/* TODO: cleaner, compiler warning friendly way to do this??? */
ws->allocStart = (BYTE*)ws->allocStart - ((size_t)ws->allocStart & (sizeof(U32)-1));
if (ws->allocStart < ws->tableValidEnd) {
ws->tableValidEnd = ws->allocStart;
}
}
ws->phase = phase;
}
}
/**
* Returns whether this object/buffer/etc was allocated in this workspace.
*/
MEM_STATIC int ZSTD_cwksp_owns_buffer(const ZSTD_cwksp* ws, const void* ptr) {
return (ptr != NULL) && (ws->workspace <= ptr) && (ptr <= ws->workspaceEnd);
}
/**
* Internal function. Do not use directly.
*/
MEM_STATIC void* ZSTD_cwksp_reserve_internal(
ZSTD_cwksp* ws, size_t bytes, ZSTD_cwksp_alloc_phase_e phase) {
void* alloc;
void* bottom = ws->tableEnd;
ZSTD_cwksp_internal_advance_phase(ws, phase);
alloc = (BYTE *)ws->allocStart - bytes;
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
/* over-reserve space */
alloc = (BYTE *)alloc - 2 * ZSTD_CWKSP_ASAN_REDZONE_SIZE;
#endif
DEBUGLOG(5, "cwksp: reserving %p %zd bytes, %zd bytes remaining",
alloc, bytes, ZSTD_cwksp_available_space(ws) - bytes);
ZSTD_cwksp_assert_internal_consistency(ws);
assert(alloc >= bottom);
if (alloc < bottom) {
DEBUGLOG(4, "cwksp: alloc failed!");
ws->allocFailed = 1;
return NULL;
}
if (alloc < ws->tableValidEnd) {
ws->tableValidEnd = alloc;
}
ws->allocStart = alloc;
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
/* Move alloc so there's ZSTD_CWKSP_ASAN_REDZONE_SIZE unused space on
* either size. */
alloc = (BYTE *)alloc + ZSTD_CWKSP_ASAN_REDZONE_SIZE;
__asan_unpoison_memory_region(alloc, bytes);
#endif
return alloc;
}
/**
* Reserves and returns unaligned memory.
*/
MEM_STATIC BYTE* ZSTD_cwksp_reserve_buffer(ZSTD_cwksp* ws, size_t bytes) {
return (BYTE*)ZSTD_cwksp_reserve_internal(ws, bytes, ZSTD_cwksp_alloc_buffers);
}
/**
* Reserves and returns memory sized on and aligned on sizeof(unsigned).
*/
MEM_STATIC void* ZSTD_cwksp_reserve_aligned(ZSTD_cwksp* ws, size_t bytes) {
assert((bytes & (sizeof(U32)-1)) == 0);
return ZSTD_cwksp_reserve_internal(ws, ZSTD_cwksp_align(bytes, sizeof(U32)), ZSTD_cwksp_alloc_aligned);
}
/**
* Aligned on sizeof(unsigned). These buffers have the special property that
* their values remain constrained, allowing us to re-use them without
* memset()-ing them.
*/
MEM_STATIC void* ZSTD_cwksp_reserve_table(ZSTD_cwksp* ws, size_t bytes) {
const ZSTD_cwksp_alloc_phase_e phase = ZSTD_cwksp_alloc_aligned;
void* alloc = ws->tableEnd;
void* end = (BYTE *)alloc + bytes;
void* top = ws->allocStart;
DEBUGLOG(5, "cwksp: reserving %p table %zd bytes, %zd bytes remaining",
alloc, bytes, ZSTD_cwksp_available_space(ws) - bytes);
assert((bytes & (sizeof(U32)-1)) == 0);
ZSTD_cwksp_internal_advance_phase(ws, phase);
ZSTD_cwksp_assert_internal_consistency(ws);
assert(end <= top);
if (end > top) {
DEBUGLOG(4, "cwksp: table alloc failed!");
ws->allocFailed = 1;
return NULL;
}
ws->tableEnd = end;
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
__asan_unpoison_memory_region(alloc, bytes);
#endif
return alloc;
}
/**
* Aligned on sizeof(void*).
*/
MEM_STATIC void* ZSTD_cwksp_reserve_object(ZSTD_cwksp* ws, size_t bytes) {
size_t roundedBytes = ZSTD_cwksp_align(bytes, sizeof(void*));
void* alloc = ws->objectEnd;
void* end = (BYTE*)alloc + roundedBytes;
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
/* over-reserve space */
end = (BYTE *)end + 2 * ZSTD_CWKSP_ASAN_REDZONE_SIZE;
#endif
DEBUGLOG(5,
"cwksp: reserving %p object %zd bytes (rounded to %zd), %zd bytes remaining",
alloc, bytes, roundedBytes, ZSTD_cwksp_available_space(ws) - roundedBytes);
assert(((size_t)alloc & (sizeof(void*)-1)) == 0);
assert((bytes & (sizeof(void*)-1)) == 0);
ZSTD_cwksp_assert_internal_consistency(ws);
/* we must be in the first phase, no advance is possible */
if (ws->phase != ZSTD_cwksp_alloc_objects || end > ws->workspaceEnd) {
DEBUGLOG(4, "cwksp: object alloc failed!");
ws->allocFailed = 1;
return NULL;
}
ws->objectEnd = end;
ws->tableEnd = end;
ws->tableValidEnd = end;
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
/* Move alloc so there's ZSTD_CWKSP_ASAN_REDZONE_SIZE unused space on
* either size. */
alloc = (BYTE *)alloc + ZSTD_CWKSP_ASAN_REDZONE_SIZE;
__asan_unpoison_memory_region(alloc, bytes);
#endif
return alloc;
}
MEM_STATIC void ZSTD_cwksp_mark_tables_dirty(ZSTD_cwksp* ws) {
DEBUGLOG(4, "cwksp: ZSTD_cwksp_mark_tables_dirty");
#if defined (MEMORY_SANITIZER) && !defined (ZSTD_MSAN_DONT_POISON_WORKSPACE)
/* To validate that the table re-use logic is sound, and that we don't
* access table space that we haven't cleaned, we re-"poison" the table
* space every time we mark it dirty. */
{
size_t size = (BYTE*)ws->tableValidEnd - (BYTE*)ws->objectEnd;
assert(__msan_test_shadow(ws->objectEnd, size) == -1);
__msan_poison(ws->objectEnd, size);
}
#endif
assert(ws->tableValidEnd >= ws->objectEnd);
assert(ws->tableValidEnd <= ws->allocStart);
ws->tableValidEnd = ws->objectEnd;
ZSTD_cwksp_assert_internal_consistency(ws);
}
MEM_STATIC void ZSTD_cwksp_mark_tables_clean(ZSTD_cwksp* ws) {
DEBUGLOG(4, "cwksp: ZSTD_cwksp_mark_tables_clean");
assert(ws->tableValidEnd >= ws->objectEnd);
assert(ws->tableValidEnd <= ws->allocStart);
if (ws->tableValidEnd < ws->tableEnd) {
ws->tableValidEnd = ws->tableEnd;
}
ZSTD_cwksp_assert_internal_consistency(ws);
}
/**
* Zero the part of the allocated tables not already marked clean.
*/
MEM_STATIC void ZSTD_cwksp_clean_tables(ZSTD_cwksp* ws) {
DEBUGLOG(4, "cwksp: ZSTD_cwksp_clean_tables");
assert(ws->tableValidEnd >= ws->objectEnd);
assert(ws->tableValidEnd <= ws->allocStart);
if (ws->tableValidEnd < ws->tableEnd) {
memset(ws->tableValidEnd, 0, (BYTE*)ws->tableEnd - (BYTE*)ws->tableValidEnd);
}
ZSTD_cwksp_mark_tables_clean(ws);
}
/**
* Invalidates table allocations.
* All other allocations remain valid.
*/
MEM_STATIC void ZSTD_cwksp_clear_tables(ZSTD_cwksp* ws) {
DEBUGLOG(4, "cwksp: clearing tables!");
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
{
size_t size = (BYTE*)ws->tableValidEnd - (BYTE*)ws->objectEnd;
__asan_poison_memory_region(ws->objectEnd, size);
}
#endif
ws->tableEnd = ws->objectEnd;
ZSTD_cwksp_assert_internal_consistency(ws);
}
/**
* Invalidates all buffer, aligned, and table allocations.
* Object allocations remain valid.
*/
MEM_STATIC void ZSTD_cwksp_clear(ZSTD_cwksp* ws) {
DEBUGLOG(4, "cwksp: clearing!");
#if defined (MEMORY_SANITIZER) && !defined (ZSTD_MSAN_DONT_POISON_WORKSPACE)
/* To validate that the context re-use logic is sound, and that we don't
* access stuff that this compression hasn't initialized, we re-"poison"
* the workspace (or at least the non-static, non-table parts of it)
* every time we start a new compression. */
{
size_t size = (BYTE*)ws->workspaceEnd - (BYTE*)ws->tableValidEnd;
__msan_poison(ws->tableValidEnd, size);
}
#endif
#if defined (ADDRESS_SANITIZER) && !defined (ZSTD_ASAN_DONT_POISON_WORKSPACE)
{
size_t size = (BYTE*)ws->workspaceEnd - (BYTE*)ws->objectEnd;
__asan_poison_memory_region(ws->objectEnd, size);
}
#endif
ws->tableEnd = ws->objectEnd;
ws->allocStart = ws->workspaceEnd;
ws->allocFailed = 0;
if (ws->phase > ZSTD_cwksp_alloc_buffers) {
ws->phase = ZSTD_cwksp_alloc_buffers;
}
ZSTD_cwksp_assert_internal_consistency(ws);
}
/**
* The provided workspace takes ownership of the buffer [start, start+size).
* Any existing values in the workspace are ignored (the previously managed
* buffer, if present, must be separately freed).
*/
MEM_STATIC void ZSTD_cwksp_init(ZSTD_cwksp* ws, void* start, size_t size) {
DEBUGLOG(4, "cwksp: init'ing workspace with %zd bytes", size);
assert(((size_t)start & (sizeof(void*)-1)) == 0); /* ensure correct alignment */
ws->workspace = start;
ws->workspaceEnd = (BYTE*)start + size;
ws->objectEnd = ws->workspace;
ws->tableValidEnd = ws->objectEnd;
ws->phase = ZSTD_cwksp_alloc_objects;
ZSTD_cwksp_clear(ws);
ws->workspaceOversizedDuration = 0;
ZSTD_cwksp_assert_internal_consistency(ws);
}
MEM_STATIC size_t ZSTD_cwksp_create(ZSTD_cwksp* ws, size_t size, ZSTD_customMem customMem) {
void* workspace = ZSTD_malloc(size, customMem);
DEBUGLOG(4, "cwksp: creating new workspace with %zd bytes", size);
RETURN_ERROR_IF(workspace == NULL, memory_allocation, "NULL pointer!");
ZSTD_cwksp_init(ws, workspace, size);
return 0;
}
MEM_STATIC void ZSTD_cwksp_free(ZSTD_cwksp* ws, ZSTD_customMem customMem) {
void *ptr = ws->workspace;
DEBUGLOG(4, "cwksp: freeing workspace");
memset(ws, 0, sizeof(ZSTD_cwksp));
ZSTD_free(ptr, customMem);
}
/**
* Moves the management of a workspace from one cwksp to another. The src cwksp
* is left in an invalid state (src must be re-init()'ed before its used again).
*/
MEM_STATIC void ZSTD_cwksp_move(ZSTD_cwksp* dst, ZSTD_cwksp* src) {
*dst = *src;
memset(src, 0, sizeof(ZSTD_cwksp));
}
MEM_STATIC size_t ZSTD_cwksp_sizeof(const ZSTD_cwksp* ws) {
return (size_t)((BYTE*)ws->workspaceEnd - (BYTE*)ws->workspace);
}
MEM_STATIC int ZSTD_cwksp_reserve_failed(const ZSTD_cwksp* ws) {
return ws->allocFailed;
}
/*-*************************************
* Functions Checking Free Space
***************************************/
MEM_STATIC size_t ZSTD_cwksp_available_space(ZSTD_cwksp* ws) {
return (size_t)((BYTE*)ws->allocStart - (BYTE*)ws->tableEnd);
}
MEM_STATIC int ZSTD_cwksp_check_available(ZSTD_cwksp* ws, size_t additionalNeededSpace) {
return ZSTD_cwksp_available_space(ws) >= additionalNeededSpace;
}
MEM_STATIC int ZSTD_cwksp_check_too_large(ZSTD_cwksp* ws, size_t additionalNeededSpace) {
return ZSTD_cwksp_check_available(
ws, additionalNeededSpace * ZSTD_WORKSPACETOOLARGE_FACTOR);
}
MEM_STATIC int ZSTD_cwksp_check_wasteful(ZSTD_cwksp* ws, size_t additionalNeededSpace) {
return ZSTD_cwksp_check_too_large(ws, additionalNeededSpace)
&& ws->workspaceOversizedDuration > ZSTD_WORKSPACETOOLARGE_MAXDURATION;
}
MEM_STATIC void ZSTD_cwksp_bump_oversized_duration(
ZSTD_cwksp* ws, size_t additionalNeededSpace) {
if (ZSTD_cwksp_check_too_large(ws, additionalNeededSpace)) {
ws->workspaceOversizedDuration++;
} else {
ws->workspaceOversizedDuration = 0;
}
}
#if defined (__cplusplus)
}
#endif
#endif /* ZSTD_CWKSP_H */

244
drivers/filesystems/btrfs/zstd/zstd_ddict.c Normal file
View file

@ -0,0 +1,244 @@
/*
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
/* zstd_ddict.c :
* concentrates all logic that needs to know the internals of ZSTD_DDict object */
/*-*******************************************************
* Dependencies
*********************************************************/
#include <string.h> /* memcpy, memmove, memset */
#include "cpu.h" /* bmi2 */
#include "mem.h" /* low level memory routines */
#define FSE_STATIC_LINKING_ONLY
#include "fse.h"
#define HUF_STATIC_LINKING_ONLY
#include "huf.h"
#include "zstd_decompress_internal.h"
#include "zstd_ddict.h"
#if defined(ZSTD_LEGACY_SUPPORT) && (ZSTD_LEGACY_SUPPORT>=1)
# include "../legacy/zstd_legacy.h"
#endif
/*-*******************************************************
* Types
*********************************************************/
struct ZSTD_DDict_s {
void* dictBuffer;
const void* dictContent;
size_t dictSize;
ZSTD_entropyDTables_t entropy;
U32 dictID;
U32 entropyPresent;
ZSTD_customMem cMem;
}; /* typedef'd to ZSTD_DDict within "zstd.h" */
const void* ZSTD_DDict_dictContent(const ZSTD_DDict* ddict)
{
assert(ddict != NULL);
return ddict->dictContent;
}
size_t ZSTD_DDict_dictSize(const ZSTD_DDict* ddict)
{
assert(ddict != NULL);
return ddict->dictSize;
}
void ZSTD_copyDDictParameters(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict)
{
DEBUGLOG(4, "ZSTD_copyDDictParameters");
assert(dctx != NULL);
assert(ddict != NULL);
dctx->dictID = ddict->dictID;
dctx->prefixStart = ddict->dictContent;
dctx->virtualStart = ddict->dictContent;
dctx->dictEnd = (const BYTE*)ddict->dictContent + ddict->dictSize;
dctx->previousDstEnd = dctx->dictEnd;
#ifdef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
dctx->dictContentBeginForFuzzing = dctx->prefixStart;
dctx->dictContentEndForFuzzing = dctx->previousDstEnd;
#endif
if (ddict->entropyPresent) {
dctx->litEntropy = 1;
dctx->fseEntropy = 1;
dctx->LLTptr = ddict->entropy.LLTable;
dctx->MLTptr = ddict->entropy.MLTable;
dctx->OFTptr = ddict->entropy.OFTable;
dctx->HUFptr = ddict->entropy.hufTable;
dctx->entropy.rep[0] = ddict->entropy.rep[0];
dctx->entropy.rep[1] = ddict->entropy.rep[1];
dctx->entropy.rep[2] = ddict->entropy.rep[2];
} else {
dctx->litEntropy = 0;
dctx->fseEntropy = 0;
}
}
static size_t
ZSTD_loadEntropy_intoDDict(ZSTD_DDict* ddict,
ZSTD_dictContentType_e dictContentType)
{
ddict->dictID = 0;
ddict->entropyPresent = 0;
if (dictContentType == ZSTD_dct_rawContent) return 0;
if (ddict->dictSize < 8) {
if (dictContentType == ZSTD_dct_fullDict)
return ERROR(dictionary_corrupted); /* only accept specified dictionaries */
return 0; /* pure content mode */
}
{ U32 const magic = MEM_readLE32(ddict->dictContent);
if (magic != ZSTD_MAGIC_DICTIONARY) {
if (dictContentType == ZSTD_dct_fullDict)
return ERROR(dictionary_corrupted); /* only accept specified dictionaries */
return 0; /* pure content mode */
}
}
ddict->dictID = MEM_readLE32((const char*)ddict->dictContent + ZSTD_FRAMEIDSIZE);
/* load entropy tables */
RETURN_ERROR_IF(ZSTD_isError(ZSTD_loadDEntropy(
&ddict->entropy, ddict->dictContent, ddict->dictSize)),
dictionary_corrupted, "");
ddict->entropyPresent = 1;
return 0;
}
static size_t ZSTD_initDDict_internal(ZSTD_DDict* ddict,
const void* dict, size_t dictSize,
ZSTD_dictLoadMethod_e dictLoadMethod,
ZSTD_dictContentType_e dictContentType)
{
if ((dictLoadMethod == ZSTD_dlm_byRef) || (!dict) || (!dictSize)) {
ddict->dictBuffer = NULL;
ddict->dictContent = dict;
if (!dict) dictSize = 0;
} else {
void* const internalBuffer = ZSTD_malloc(dictSize, ddict->cMem);
ddict->dictBuffer = internalBuffer;
ddict->dictContent = internalBuffer;
if (!internalBuffer) return ERROR(memory_allocation);
memcpy(internalBuffer, dict, dictSize);
}
ddict->dictSize = dictSize;
ddict->entropy.hufTable[0] = (HUF_DTable)((HufLog)*0x1000001); /* cover both little and big endian */
/* parse dictionary content */
FORWARD_IF_ERROR( ZSTD_loadEntropy_intoDDict(ddict, dictContentType) , "");
return 0;
}
ZSTD_DDict* ZSTD_createDDict_advanced(const void* dict, size_t dictSize,
ZSTD_dictLoadMethod_e dictLoadMethod,
ZSTD_dictContentType_e dictContentType,
ZSTD_customMem customMem)
{
if (!customMem.customAlloc ^ !customMem.customFree) return NULL;
{ ZSTD_DDict* const ddict = (ZSTD_DDict*) ZSTD_malloc(sizeof(ZSTD_DDict), customMem);
if (ddict == NULL) return NULL;
ddict->cMem = customMem;
{ size_t const initResult = ZSTD_initDDict_internal(ddict,
dict, dictSize,
dictLoadMethod, dictContentType);
if (ZSTD_isError(initResult)) {
ZSTD_freeDDict(ddict);
return NULL;
} }
return ddict;
}
}
/*! ZSTD_createDDict() :
* Create a digested dictionary, to start decompression without startup delay.
* `dict` content is copied inside DDict.
* Consequently, `dict` can be released after `ZSTD_DDict` creation */
ZSTD_DDict* ZSTD_createDDict(const void* dict, size_t dictSize)
{
ZSTD_customMem const allocator = { NULL, NULL, NULL };
return ZSTD_createDDict_advanced(dict, dictSize, ZSTD_dlm_byCopy, ZSTD_dct_auto, allocator);
}
/*! ZSTD_createDDict_byReference() :
* Create a digested dictionary, to start decompression without startup delay.
* Dictionary content is simply referenced, it will be accessed during decompression.
* Warning : dictBuffer must outlive DDict (DDict must be freed before dictBuffer) */
ZSTD_DDict* ZSTD_createDDict_byReference(const void* dictBuffer, size_t dictSize)
{
ZSTD_customMem const allocator = { NULL, NULL, NULL };
return ZSTD_createDDict_advanced(dictBuffer, dictSize, ZSTD_dlm_byRef, ZSTD_dct_auto, allocator);
}
const ZSTD_DDict* ZSTD_initStaticDDict(
void* sBuffer, size_t sBufferSize,
const void* dict, size_t dictSize,
ZSTD_dictLoadMethod_e dictLoadMethod,
ZSTD_dictContentType_e dictContentType)
{
size_t const neededSpace = sizeof(ZSTD_DDict)
+ (dictLoadMethod == ZSTD_dlm_byRef ? 0 : dictSize);
ZSTD_DDict* const ddict = (ZSTD_DDict*)sBuffer;
assert(sBuffer != NULL);
assert(dict != NULL);
if ((size_t)sBuffer & 7) return NULL; /* 8-aligned */
if (sBufferSize < neededSpace) return NULL;
if (dictLoadMethod == ZSTD_dlm_byCopy) {
memcpy(ddict+1, dict, dictSize); /* local copy */
dict = ddict+1;
}
if (ZSTD_isError( ZSTD_initDDict_internal(ddict,
dict, dictSize,
ZSTD_dlm_byRef, dictContentType) ))
return NULL;
return ddict;
}
size_t ZSTD_freeDDict(ZSTD_DDict* ddict)
{
if (ddict==NULL) return 0; /* support free on NULL */
{ ZSTD_customMem const cMem = ddict->cMem;
ZSTD_free(ddict->dictBuffer, cMem);
ZSTD_free(ddict, cMem);
return 0;
}
}
/*! ZSTD_estimateDDictSize() :
* Estimate amount of memory that will be needed to create a dictionary for decompression.
* Note : dictionary created by reference using ZSTD_dlm_byRef are smaller */
size_t ZSTD_estimateDDictSize(size_t dictSize, ZSTD_dictLoadMethod_e dictLoadMethod)
{
return sizeof(ZSTD_DDict) + (dictLoadMethod == ZSTD_dlm_byRef ? 0 : dictSize);
}
size_t ZSTD_sizeof_DDict(const ZSTD_DDict* ddict)
{
if (ddict==NULL) return 0; /* support sizeof on NULL */
return sizeof(*ddict) + (ddict->dictBuffer ? ddict->dictSize : 0) ;
}
/*! ZSTD_getDictID_fromDDict() :
* Provides the dictID of the dictionary loaded into `ddict`.
* If @return == 0, the dictionary is not conformant to Zstandard specification, or empty.
* Non-conformant dictionaries can still be loaded, but as content-only dictionaries. */
unsigned ZSTD_getDictID_fromDDict(const ZSTD_DDict* ddict)
{
if (ddict==NULL) return 0;
return ZSTD_getDictID_fromDict(ddict->dictContent, ddict->dictSize);
}

44
drivers/filesystems/btrfs/zstd/zstd_ddict.h Normal file
View file

@ -0,0 +1,44 @@
/*
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
#ifndef ZSTD_DDICT_H
#define ZSTD_DDICT_H
/*-*******************************************************
* Dependencies
*********************************************************/
#include <stddef.h> /* size_t */
#include "zstd.h" /* ZSTD_DDict, and several public functions */
/*-*******************************************************
* Interface
*********************************************************/
/* note: several prototypes are already published in `zstd.h` :
* ZSTD_createDDict()
* ZSTD_createDDict_byReference()
* ZSTD_createDDict_advanced()
* ZSTD_freeDDict()
* ZSTD_initStaticDDict()
* ZSTD_sizeof_DDict()
* ZSTD_estimateDDictSize()
* ZSTD_getDictID_fromDict()
*/
const void* ZSTD_DDict_dictContent(const ZSTD_DDict* ddict);
size_t ZSTD_DDict_dictSize(const ZSTD_DDict* ddict);
void ZSTD_copyDDictParameters(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict);
#endif /* ZSTD_DDICT_H */

2499
drivers/filesystems/btrfs/zstd/zstd_decompress.c
View file

File diff suppressed because it is too large Load diff

1432
drivers/filesystems/btrfs/zstd/zstd_decompress_block.c Normal file
View file

File diff suppressed because it is too large Load diff

59
drivers/filesystems/btrfs/zstd/zstd_decompress_block.h Normal file
View file

@ -0,0 +1,59 @@
/*
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
#ifndef ZSTD_DEC_BLOCK_H
#define ZSTD_DEC_BLOCK_H
/*-*******************************************************
* Dependencies
*********************************************************/
#include <stddef.h> /* size_t */
#include "zstd.h" /* DCtx, and some public functions */
#include "zstd_internal.h" /* blockProperties_t, and some public functions */
#include "zstd_decompress_internal.h" /* ZSTD_seqSymbol */
/* === Prototypes === */
/* note: prototypes already published within `zstd.h` :
* ZSTD_decompressBlock()
*/
/* note: prototypes already published within `zstd_internal.h` :
* ZSTD_getcBlockSize()
* ZSTD_decodeSeqHeaders()
*/
/* ZSTD_decompressBlock_internal() :
* decompress block, starting at `src`,
* into destination buffer `dst`.
* @return : decompressed block size,
* or an error code (which can be tested using ZSTD_isError())
*/
size_t ZSTD_decompressBlock_internal(ZSTD_DCtx* dctx,
void* dst, size_t dstCapacity,
const void* src, size_t srcSize, const int frame);
/* ZSTD_buildFSETable() :
* generate FSE decoding table for one symbol (ll, ml or off)
* this function must be called with valid parameters only
* (dt is large enough, normalizedCounter distribution total is a power of 2, max is within range, etc.)
* in which case it cannot fail.
* Internal use only.
*/
void ZSTD_buildFSETable(ZSTD_seqSymbol* dt,
const short* normalizedCounter, unsigned maxSymbolValue,
const U32* baseValue, const U32* nbAdditionalBits,
unsigned tableLog);
#endif /* ZSTD_DEC_BLOCK_H */

189
drivers/filesystems/btrfs/zstd/zstd_decompress_internal.h Normal file
View file

@ -0,0 +1,189 @@
/*
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
/* zstd_decompress_internal:
* objects and definitions shared within lib/decompress modules */
#ifndef ZSTD_DECOMPRESS_INTERNAL_H
#define ZSTD_DECOMPRESS_INTERNAL_H
/*-*******************************************************
* Dependencies
*********************************************************/
#include "mem.h" /* BYTE, U16, U32 */
#include "zstd_internal.h" /* ZSTD_seqSymbol */
/*-*******************************************************
* Constants
*********************************************************/
static const U32 LL_base[MaxLL+1] = {
0, 1, 2, 3, 4, 5, 6, 7,
8, 9, 10, 11, 12, 13, 14, 15,
16, 18, 20, 22, 24, 28, 32, 40,
48, 64, 0x80, 0x100, 0x200, 0x400, 0x800, 0x1000,
0x2000, 0x4000, 0x8000, 0x10000 };
static const U32 OF_base[MaxOff+1] = {
0, 1, 1, 5, 0xD, 0x1D, 0x3D, 0x7D,
0xFD, 0x1FD, 0x3FD, 0x7FD, 0xFFD, 0x1FFD, 0x3FFD, 0x7FFD,
0xFFFD, 0x1FFFD, 0x3FFFD, 0x7FFFD, 0xFFFFD, 0x1FFFFD, 0x3FFFFD, 0x7FFFFD,
0xFFFFFD, 0x1FFFFFD, 0x3FFFFFD, 0x7FFFFFD, 0xFFFFFFD, 0x1FFFFFFD, 0x3FFFFFFD, 0x7FFFFFFD };
static const U32 OF_bits[MaxOff+1] = {
0, 1, 2, 3, 4, 5, 6, 7,
8, 9, 10, 11, 12, 13, 14, 15,
16, 17, 18, 19, 20, 21, 22, 23,
24, 25, 26, 27, 28, 29, 30, 31 };
static const U32 ML_base[MaxML+1] = {
3, 4, 5, 6, 7, 8, 9, 10,
11, 12, 13, 14, 15, 16, 17, 18,
19, 20, 21, 22, 23, 24, 25, 26,
27, 28, 29, 30, 31, 32, 33, 34,
35, 37, 39, 41, 43, 47, 51, 59,
67, 83, 99, 0x83, 0x103, 0x203, 0x403, 0x803,
0x1003, 0x2003, 0x4003, 0x8003, 0x10003 };
/*-*******************************************************
* Decompression types
*********************************************************/
typedef struct {
U32 fastMode;
U32 tableLog;
} ZSTD_seqSymbol_header;
typedef struct {
U16 nextState;
BYTE nbAdditionalBits;
BYTE nbBits;
U32 baseValue;
} ZSTD_seqSymbol;
#define SEQSYMBOL_TABLE_SIZE(log) (1 + (1 << (log)))
typedef struct {
ZSTD_seqSymbol LLTable[SEQSYMBOL_TABLE_SIZE(LLFSELog)]; /* Note : Space reserved for FSE Tables */
ZSTD_seqSymbol OFTable[SEQSYMBOL_TABLE_SIZE(OffFSELog)]; /* is also used as temporary workspace while building hufTable during DDict creation */
ZSTD_seqSymbol MLTable[SEQSYMBOL_TABLE_SIZE(MLFSELog)]; /* and therefore must be at least HUF_DECOMPRESS_WORKSPACE_SIZE large */
HUF_DTable hufTable[HUF_DTABLE_SIZE(HufLog)]; /* can accommodate HUF_decompress4X */
U32 rep[ZSTD_REP_NUM];
} ZSTD_entropyDTables_t;
typedef enum { ZSTDds_getFrameHeaderSize, ZSTDds_decodeFrameHeader,
ZSTDds_decodeBlockHeader, ZSTDds_decompressBlock,
ZSTDds_decompressLastBlock, ZSTDds_checkChecksum,
ZSTDds_decodeSkippableHeader, ZSTDds_skipFrame } ZSTD_dStage;
typedef enum { zdss_init=0, zdss_loadHeader,
zdss_read, zdss_load, zdss_flush } ZSTD_dStreamStage;
typedef enum {
ZSTD_use_indefinitely = -1, /* Use the dictionary indefinitely */
ZSTD_dont_use = 0, /* Do not use the dictionary (if one exists free it) */
ZSTD_use_once = 1 /* Use the dictionary once and set to ZSTD_dont_use */
} ZSTD_dictUses_e;
typedef enum {
ZSTD_obm_buffered = 0, /* Buffer the output */
ZSTD_obm_stable = 1 /* ZSTD_outBuffer is stable */
} ZSTD_outBufferMode_e;
struct ZSTD_DCtx_s
{
const ZSTD_seqSymbol* LLTptr;
const ZSTD_seqSymbol* MLTptr;
const ZSTD_seqSymbol* OFTptr;
const HUF_DTable* HUFptr;
ZSTD_entropyDTables_t entropy;
U32 workspace[HUF_DECOMPRESS_WORKSPACE_SIZE_U32]; /* space needed when building huffman tables */
const void* previousDstEnd; /* detect continuity */
const void* prefixStart; /* start of current segment */
const void* virtualStart; /* virtual start of previous segment if it was just before current one */
const void* dictEnd; /* end of previous segment */
size_t expected;
ZSTD_frameHeader fParams;
U64 decodedSize;
blockType_e bType; /* used in ZSTD_decompressContinue(), store blockType between block header decoding and block decompression stages */
ZSTD_dStage stage;
U32 litEntropy;
U32 fseEntropy;
XXH64_state_t xxhState;
size_t headerSize;
ZSTD_format_e format;
const BYTE* litPtr;
ZSTD_customMem customMem;
size_t litSize;
size_t rleSize;
size_t staticSize;
int bmi2; /* == 1 if the CPU supports BMI2 and 0 otherwise. CPU support is determined dynamically once per context lifetime. */
/* dictionary */
ZSTD_DDict* ddictLocal;
const ZSTD_DDict* ddict; /* set by ZSTD_initDStream_usingDDict(), or ZSTD_DCtx_refDDict() */
U32 dictID;
int ddictIsCold; /* if == 1 : dictionary is "new" for working context, and presumed "cold" (not in cpu cache) */
ZSTD_dictUses_e dictUses;
/* streaming */
ZSTD_dStreamStage streamStage;
char* inBuff;
size_t inBuffSize;
size_t inPos;
size_t maxWindowSize;
char* outBuff;
size_t outBuffSize;
size_t outStart;
size_t outEnd;
size_t lhSize;
void* legacyContext;
U32 previousLegacyVersion;
U32 legacyVersion;
U32 hostageByte;
int noForwardProgress;
ZSTD_outBufferMode_e outBufferMode;
ZSTD_outBuffer expectedOutBuffer;
/* workspace */
BYTE litBuffer[ZSTD_BLOCKSIZE_MAX + WILDCOPY_OVERLENGTH];
BYTE headerBuffer[ZSTD_FRAMEHEADERSIZE_MAX];
size_t oversizedDuration;
#ifdef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
void const* dictContentBeginForFuzzing;
void const* dictContentEndForFuzzing;
#endif
}; /* typedef'd to ZSTD_DCtx within "zstd.h" */
/*-*******************************************************
* Shared internal functions
*********************************************************/
/*! ZSTD_loadDEntropy() :
* dict : must point at beginning of a valid zstd dictionary.
* @return : size of dictionary header (size of magic number + dict ID + entropy tables) */
size_t ZSTD_loadDEntropy(ZSTD_entropyDTables_t* entropy,
const void* const dict, size_t const dictSize);
/*! ZSTD_checkContinuity() :
* check if next `dst` follows previous position, where decompression ended.
* If yes, do nothing (continue on current segment).
* If not, classify previous segment as "external dictionary", and start a new segment.
* This function cannot fail. */
void ZSTD_checkContinuity(ZSTD_DCtx* dctx, const void* dst);
#endif /* ZSTD_DECOMPRESS_INTERNAL_H */

114
drivers/filesystems/btrfs/zstd/zstd_double_fast.c
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -18,7 +18,7 @@ void ZSTD_fillDoubleHashTable(ZSTD_matchState_t* ms,
const ZSTD_compressionParameters* const cParams = &ms->cParams;
U32* const hashLarge = ms->hashTable;
U32 const hBitsL = cParams->hashLog;
U32 const mls = cParams->searchLength;
U32 const mls = cParams->minMatch;
U32* const hashSmall = ms->chainTable;
U32 const hBitsS = cParams->chainLog;
const BYTE* const base = ms->window.base;
@ -43,8 +43,7 @@ void ZSTD_fillDoubleHashTable(ZSTD_matchState_t* ms,
/* Only load extra positions for ZSTD_dtlm_full */
if (dtlm == ZSTD_dtlm_fast)
break;
}
}
} }
}
@ -63,7 +62,9 @@ size_t ZSTD_compressBlock_doubleFast_generic(
const BYTE* const istart = (const BYTE*)src;
const BYTE* ip = istart;
const BYTE* anchor = istart;
const U32 prefixLowestIndex = ms->window.dictLimit;
const U32 endIndex = (U32)((size_t)(istart - base) + srcSize);
/* presumes that, if there is a dictionary, it must be using Attach mode */
const U32 prefixLowestIndex = ZSTD_getLowestPrefixIndex(ms, endIndex, cParams->windowLog);
const BYTE* const prefixLowest = base + prefixLowestIndex;
const BYTE* const iend = istart + srcSize;
const BYTE* const ilimit = iend - HASH_READ_SIZE;
@ -93,14 +94,23 @@ size_t ZSTD_compressBlock_doubleFast_generic(
dictCParams->hashLog : hBitsL;
const U32 dictHBitsS = dictMode == ZSTD_dictMatchState ?
dictCParams->chainLog : hBitsS;
const U32 dictAndPrefixLength = (U32)(ip - prefixLowest + dictEnd - dictStart);
const U32 dictAndPrefixLength = (U32)((ip - prefixLowest) + (dictEnd - dictStart));
DEBUGLOG(5, "ZSTD_compressBlock_doubleFast_generic");
assert(dictMode == ZSTD_noDict || dictMode == ZSTD_dictMatchState);
/* if a dictionary is attached, it must be within window range */
if (dictMode == ZSTD_dictMatchState) {
assert(ms->window.dictLimit + (1U << cParams->windowLog) >= endIndex);
}
/* init */
ip += (dictAndPrefixLength == 0);
if (dictMode == ZSTD_noDict) {
U32 const maxRep = (U32)(ip - prefixLowest);
U32 const current = (U32)(ip - base);
U32 const windowLow = ZSTD_getLowestPrefixIndex(ms, current, cParams->windowLog);
U32 const maxRep = current - windowLow;
if (offset_2 > maxRep) offsetSaved = offset_2, offset_2 = 0;
if (offset_1 > maxRep) offsetSaved = offset_1, offset_1 = 0;
}
@ -138,7 +148,7 @@ size_t ZSTD_compressBlock_doubleFast_generic(
const BYTE* repMatchEnd = repIndex < prefixLowestIndex ? dictEnd : iend;
mLength = ZSTD_count_2segments(ip+1+4, repMatch+4, iend, repMatchEnd, prefixLowest) + 4;
ip++;
ZSTD_storeSeq(seqStore, ip-anchor, anchor, 0, mLength-MINMATCH);
ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, 0, mLength-MINMATCH);
goto _match_stored;
}
@ -147,7 +157,7 @@ size_t ZSTD_compressBlock_doubleFast_generic(
&& ((offset_1 > 0) & (MEM_read32(ip+1-offset_1) == MEM_read32(ip+1)))) {
mLength = ZSTD_count(ip+1+4, ip+1+4-offset_1, iend) + 4;
ip++;
ZSTD_storeSeq(seqStore, ip-anchor, anchor, 0, mLength-MINMATCH);
ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, 0, mLength-MINMATCH);
goto _match_stored;
}
@ -170,8 +180,7 @@ size_t ZSTD_compressBlock_doubleFast_generic(
offset = (U32)(current - dictMatchIndexL - dictIndexDelta);
while (((ip>anchor) & (dictMatchL>dictStart)) && (ip[-1] == dictMatchL[-1])) { ip--; dictMatchL--; mLength++; } /* catch up */
goto _match_found;
}
}
} }
if (matchIndexS > prefixLowestIndex) {
/* check prefix short match */
@ -186,16 +195,17 @@ size_t ZSTD_compressBlock_doubleFast_generic(
if (match > dictStart && MEM_read32(match) == MEM_read32(ip)) {
goto _search_next_long;
}
}
} }
ip += ((ip-anchor) >> kSearchStrength) + 1;
#if defined(__aarch64__)
PREFETCH_L1(ip+256);
#endif
continue;
_search_next_long:
{
size_t const hl3 = ZSTD_hashPtr(ip+1, hBitsL, 8);
{ size_t const hl3 = ZSTD_hashPtr(ip+1, hBitsL, 8);
size_t const dictHLNext = ZSTD_hashPtr(ip+1, dictHBitsL, 8);
U32 const matchIndexL3 = hashLong[hl3];
const BYTE* matchL3 = base + matchIndexL3;
@ -221,9 +231,7 @@ _search_next_long:
offset = (U32)(current + 1 - dictMatchIndexL3 - dictIndexDelta);
while (((ip>anchor) & (dictMatchL3>dictStart)) && (ip[-1] == dictMatchL3[-1])) { ip--; dictMatchL3--; mLength++; } /* catch up */
goto _match_found;
}
}
}
} } }
/* if no long +1 match, explore the short match we found */
if (dictMode == ZSTD_dictMatchState && matchIndexS < prefixLowestIndex) {
@ -242,7 +250,7 @@ _match_found:
offset_2 = offset_1;
offset_1 = offset;
ZSTD_storeSeq(seqStore, ip-anchor, anchor, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
_match_stored:
/* match found */
@ -250,11 +258,14 @@ _match_stored:
anchor = ip;
if (ip <= ilimit) {
/* Fill Table */
hashLong[ZSTD_hashPtr(base+current+2, hBitsL, 8)] =
hashSmall[ZSTD_hashPtr(base+current+2, hBitsS, mls)] = current+2; /* here because current+2 could be > iend-8 */
hashLong[ZSTD_hashPtr(ip-2, hBitsL, 8)] =
hashSmall[ZSTD_hashPtr(ip-2, hBitsS, mls)] = (U32)(ip-2-base);
/* Complementary insertion */
/* done after iLimit test, as candidates could be > iend-8 */
{ U32 const indexToInsert = current+2;
hashLong[ZSTD_hashPtr(base+indexToInsert, hBitsL, 8)] = indexToInsert;
hashLong[ZSTD_hashPtr(ip-2, hBitsL, 8)] = (U32)(ip-2-base);
hashSmall[ZSTD_hashPtr(base+indexToInsert, hBitsS, mls)] = indexToInsert;
hashSmall[ZSTD_hashPtr(ip-1, hBitsS, mls)] = (U32)(ip-1-base);
}
/* check immediate repcode */
if (dictMode == ZSTD_dictMatchState) {
@ -263,14 +274,14 @@ _match_stored:
U32 const repIndex2 = current2 - offset_2;
const BYTE* repMatch2 = dictMode == ZSTD_dictMatchState
&& repIndex2 < prefixLowestIndex ?
dictBase - dictIndexDelta + repIndex2 :
dictBase + repIndex2 - dictIndexDelta :
base + repIndex2;
if ( ((U32)((prefixLowestIndex-1) - (U32)repIndex2) >= 3 /* intentional overflow */)
&& (MEM_read32(repMatch2) == MEM_read32(ip)) ) {
const BYTE* const repEnd2 = repIndex2 < prefixLowestIndex ? dictEnd : iend;
size_t const repLength2 = ZSTD_count_2segments(ip+4, repMatch2+4, iend, repEnd2, prefixLowest) + 4;
U32 tmpOffset = offset_2; offset_2 = offset_1; offset_1 = tmpOffset; /* swap offset_2 <=> offset_1 */
ZSTD_storeSeq(seqStore, 0, anchor, 0, repLength2-MINMATCH);
ZSTD_storeSeq(seqStore, 0, anchor, iend, 0, repLength2-MINMATCH);
hashSmall[ZSTD_hashPtr(ip, hBitsS, mls)] = current2;
hashLong[ZSTD_hashPtr(ip, hBitsL, 8)] = current2;
ip += repLength2;
@ -278,8 +289,7 @@ _match_stored:
continue;
}
break;
}
}
} }
if (dictMode == ZSTD_noDict) {
while ( (ip <= ilimit)
@ -290,18 +300,19 @@ _match_stored:
U32 const tmpOff = offset_2; offset_2 = offset_1; offset_1 = tmpOff; /* swap offset_2 <=> offset_1 */
hashSmall[ZSTD_hashPtr(ip, hBitsS, mls)] = (U32)(ip-base);
hashLong[ZSTD_hashPtr(ip, hBitsL, 8)] = (U32)(ip-base);
ZSTD_storeSeq(seqStore, 0, anchor, 0, rLength-MINMATCH);
ZSTD_storeSeq(seqStore, 0, anchor, iend, 0, rLength-MINMATCH);
ip += rLength;
anchor = ip;
continue; /* faster when present ... (?) */
} } } }
} } }
} /* while (ip < ilimit) */
/* save reps for next block */
rep[0] = offset_1 ? offset_1 : offsetSaved;
rep[1] = offset_2 ? offset_2 : offsetSaved;
/* Return the last literals size */
return iend - anchor;
return (size_t)(iend - anchor);
}
@ -309,7 +320,7 @@ size_t ZSTD_compressBlock_doubleFast(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
const U32 mls = ms->cParams.searchLength;
const U32 mls = ms->cParams.minMatch;
switch(mls)
{
default: /* includes case 3 */
@ -329,7 +340,7 @@ size_t ZSTD_compressBlock_doubleFast_dictMatchState(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
const U32 mls = ms->cParams.searchLength;
const U32 mls = ms->cParams.minMatch;
switch(mls)
{
default: /* includes case 3 */
@ -360,10 +371,13 @@ static size_t ZSTD_compressBlock_doubleFast_extDict_generic(
const BYTE* anchor = istart;
const BYTE* const iend = istart + srcSize;
const BYTE* const ilimit = iend - 8;
const U32 prefixStartIndex = ms->window.dictLimit;
const BYTE* const base = ms->window.base;
const U32 endIndex = (U32)((size_t)(istart - base) + srcSize);
const U32 lowLimit = ZSTD_getLowestMatchIndex(ms, endIndex, cParams->windowLog);
const U32 dictStartIndex = lowLimit;
const U32 dictLimit = ms->window.dictLimit;
const U32 prefixStartIndex = (dictLimit > lowLimit) ? dictLimit : lowLimit;
const BYTE* const prefixStart = base + prefixStartIndex;
const U32 dictStartIndex = ms->window.lowLimit;
const BYTE* const dictBase = ms->window.dictBase;
const BYTE* const dictStart = dictBase + dictStartIndex;
const BYTE* const dictEnd = dictBase + prefixStartIndex;
@ -371,6 +385,10 @@ static size_t ZSTD_compressBlock_doubleFast_extDict_generic(
DEBUGLOG(5, "ZSTD_compressBlock_doubleFast_extDict_generic (srcSize=%zu)", srcSize);
/* if extDict is invalidated due to maxDistance, switch to "regular" variant */
if (prefixStartIndex == dictStartIndex)
return ZSTD_compressBlock_doubleFast_generic(ms, seqStore, rep, src, srcSize, mls, ZSTD_noDict);
/* Search Loop */
while (ip < ilimit) { /* < instead of <=, because (ip+1) */
const size_t hSmall = ZSTD_hashPtr(ip, hBitsS, mls);
@ -396,7 +414,7 @@ static size_t ZSTD_compressBlock_doubleFast_extDict_generic(
const BYTE* repMatchEnd = repIndex < prefixStartIndex ? dictEnd : iend;
mLength = ZSTD_count_2segments(ip+1+4, repMatch+4, iend, repMatchEnd, prefixStart) + 4;
ip++;
ZSTD_storeSeq(seqStore, ip-anchor, anchor, 0, mLength-MINMATCH);
ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, 0, mLength-MINMATCH);
} else {
if ((matchLongIndex > dictStartIndex) && (MEM_read64(matchLong) == MEM_read64(ip))) {
const BYTE* const matchEnd = matchLongIndex < prefixStartIndex ? dictEnd : iend;
@ -407,7 +425,7 @@ static size_t ZSTD_compressBlock_doubleFast_extDict_generic(
while (((ip>anchor) & (matchLong>lowMatchPtr)) && (ip[-1] == matchLong[-1])) { ip--; matchLong--; mLength++; } /* catch up */
offset_2 = offset_1;
offset_1 = offset;
ZSTD_storeSeq(seqStore, ip-anchor, anchor, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
} else if ((matchIndex > dictStartIndex) && (MEM_read32(match) == MEM_read32(ip))) {
size_t const h3 = ZSTD_hashPtr(ip+1, hBitsL, 8);
@ -432,23 +450,27 @@ static size_t ZSTD_compressBlock_doubleFast_extDict_generic(
}
offset_2 = offset_1;
offset_1 = offset;
ZSTD_storeSeq(seqStore, ip-anchor, anchor, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
} else {
ip += ((ip-anchor) >> kSearchStrength) + 1;
continue;
} }
/* found a match : store it */
/* move to next sequence start */
ip += mLength;
anchor = ip;
if (ip <= ilimit) {
/* Fill Table */
hashSmall[ZSTD_hashPtr(base+current+2, hBitsS, mls)] = current+2;
hashLong[ZSTD_hashPtr(base+current+2, hBitsL, 8)] = current+2;
hashSmall[ZSTD_hashPtr(ip-2, hBitsS, mls)] = (U32)(ip-2-base);
/* Complementary insertion */
/* done after iLimit test, as candidates could be > iend-8 */
{ U32 const indexToInsert = current+2;
hashLong[ZSTD_hashPtr(base+indexToInsert, hBitsL, 8)] = indexToInsert;
hashLong[ZSTD_hashPtr(ip-2, hBitsL, 8)] = (U32)(ip-2-base);
hashSmall[ZSTD_hashPtr(base+indexToInsert, hBitsS, mls)] = indexToInsert;
hashSmall[ZSTD_hashPtr(ip-1, hBitsS, mls)] = (U32)(ip-1-base);
}
/* check immediate repcode */
while (ip <= ilimit) {
U32 const current2 = (U32)(ip-base);
@ -460,7 +482,7 @@ static size_t ZSTD_compressBlock_doubleFast_extDict_generic(
const BYTE* const repEnd2 = repIndex2 < prefixStartIndex ? dictEnd : iend;
size_t const repLength2 = ZSTD_count_2segments(ip+4, repMatch2+4, iend, repEnd2, prefixStart) + 4;
U32 const tmpOffset = offset_2; offset_2 = offset_1; offset_1 = tmpOffset; /* swap offset_2 <=> offset_1 */
ZSTD_storeSeq(seqStore, 0, anchor, 0, repLength2-MINMATCH);
ZSTD_storeSeq(seqStore, 0, anchor, iend, 0, repLength2-MINMATCH);
hashSmall[ZSTD_hashPtr(ip, hBitsS, mls)] = current2;
hashLong[ZSTD_hashPtr(ip, hBitsL, 8)] = current2;
ip += repLength2;
@ -475,7 +497,7 @@ static size_t ZSTD_compressBlock_doubleFast_extDict_generic(
rep[1] = offset_2;
/* Return the last literals size */
return iend - anchor;
return (size_t)(iend - anchor);
}
@ -483,7 +505,7 @@ size_t ZSTD_compressBlock_doubleFast_extDict(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
U32 const mls = ms->cParams.searchLength;
U32 const mls = ms->cParams.minMatch;
switch(mls)
{
default: /* includes case 3 */

2
drivers/filesystems/btrfs/zstd/zstd_double_fast.h
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

4
drivers/filesystems/btrfs/zstd/zstd_errors.h
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -72,9 +72,11 @@ typedef enum {
ZSTD_error_workSpace_tooSmall= 66,
ZSTD_error_dstSize_tooSmall = 70,
ZSTD_error_srcSize_wrong = 72,
ZSTD_error_dstBuffer_null = 74,
/* following error codes are __NOT STABLE__, they can be removed or changed in future versions */
ZSTD_error_frameIndex_tooLarge = 100,
ZSTD_error_seekableIO = 102,
ZSTD_error_dstBuffer_wrong = 104,
ZSTD_error_maxCode = 120 /* never EVER use this value directly, it can change in future versions! Use ZSTD_isError() instead */
} ZSTD_ErrorCode;

367
drivers/filesystems/btrfs/zstd/zstd_fast.c
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -8,17 +8,18 @@
* You may select, at your option, one of the above-listed licenses.
*/
#include "zstd_compress_internal.h"
#include "zstd_compress_internal.h" /* ZSTD_hashPtr, ZSTD_count, ZSTD_storeSeq */
#include "zstd_fast.h"
void ZSTD_fillHashTable(ZSTD_matchState_t* ms,
void const* end, ZSTD_dictTableLoadMethod_e dtlm)
const void* const end,
ZSTD_dictTableLoadMethod_e dtlm)
{
const ZSTD_compressionParameters* const cParams = &ms->cParams;
U32* const hashTable = ms->hashTable;
U32 const hBits = cParams->hashLog;
U32 const mls = cParams->searchLength;
U32 const mls = cParams->minMatch;
const BYTE* const base = ms->window.base;
const BYTE* ip = base + ms->nextToUpdate;
const BYTE* const iend = ((const BYTE*)end) - HASH_READ_SIZE;
@ -27,25 +28,185 @@ void ZSTD_fillHashTable(ZSTD_matchState_t* ms,
/* Always insert every fastHashFillStep position into the hash table.
* Insert the other positions if their hash entry is empty.
*/
for (; ip + fastHashFillStep - 1 <= iend; ip += fastHashFillStep) {
for ( ; ip + fastHashFillStep < iend + 2; ip += fastHashFillStep) {
U32 const current = (U32)(ip - base);
U32 i;
for (i = 0; i < fastHashFillStep; ++i) {
size_t const hash = ZSTD_hashPtr(ip + i, hBits, mls);
if (i == 0 || hashTable[hash] == 0)
hashTable[hash] = current + i;
size_t const hash0 = ZSTD_hashPtr(ip, hBits, mls);
hashTable[hash0] = current;
if (dtlm == ZSTD_dtlm_fast) continue;
/* Only load extra positions for ZSTD_dtlm_full */
if (dtlm == ZSTD_dtlm_fast)
break;
{ U32 p;
for (p = 1; p < fastHashFillStep; ++p) {
size_t const hash = ZSTD_hashPtr(ip + p, hBits, mls);
if (hashTable[hash] == 0) { /* not yet filled */
hashTable[hash] = current + p;
} } } }
}
FORCE_INLINE_TEMPLATE size_t
ZSTD_compressBlock_fast_generic(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize,
U32 const mls)
{
const ZSTD_compressionParameters* const cParams = &ms->cParams;
U32* const hashTable = ms->hashTable;
U32 const hlog = cParams->hashLog;
/* support stepSize of 0 */
size_t const stepSize = cParams->targetLength + !(cParams->targetLength) + 1;
const BYTE* const base = ms->window.base;
const BYTE* const istart = (const BYTE*)src;
/* We check ip0 (ip + 0) and ip1 (ip + 1) each loop */
const BYTE* ip0 = istart;
const BYTE* ip1;
const BYTE* anchor = istart;
const U32 endIndex = (U32)((size_t)(istart - base) + srcSize);
const U32 prefixStartIndex = ZSTD_getLowestPrefixIndex(ms, endIndex, cParams->windowLog);
const BYTE* const prefixStart = base + prefixStartIndex;
const BYTE* const iend = istart + srcSize;
const BYTE* const ilimit = iend - HASH_READ_SIZE;
U32 offset_1=rep[0], offset_2=rep[1];
U32 offsetSaved = 0;
/* init */
DEBUGLOG(5, "ZSTD_compressBlock_fast_generic");
ip0 += (ip0 == prefixStart);
ip1 = ip0 + 1;
{ U32 const current = (U32)(ip0 - base);
U32 const windowLow = ZSTD_getLowestPrefixIndex(ms, current, cParams->windowLog);
U32 const maxRep = current - windowLow;
if (offset_2 > maxRep) offsetSaved = offset_2, offset_2 = 0;
if (offset_1 > maxRep) offsetSaved = offset_1, offset_1 = 0;
}
/* Main Search Loop */
#ifdef __INTEL_COMPILER
/* From intel 'The vector pragma indicates that the loop should be
* vectorized if it is legal to do so'. Can be used together with
* #pragma ivdep (but have opted to exclude that because intel
* warns against using it).*/
#pragma vector always
#endif
while (ip1 < ilimit) { /* < instead of <=, because check at ip0+2 */
size_t mLength;
BYTE const* ip2 = ip0 + 2;
size_t const h0 = ZSTD_hashPtr(ip0, hlog, mls);
U32 const val0 = MEM_read32(ip0);
size_t const h1 = ZSTD_hashPtr(ip1, hlog, mls);
U32 const val1 = MEM_read32(ip1);
U32 const current0 = (U32)(ip0-base);
U32 const current1 = (U32)(ip1-base);
U32 const matchIndex0 = hashTable[h0];
U32 const matchIndex1 = hashTable[h1];
BYTE const* repMatch = ip2 - offset_1;
const BYTE* match0 = base + matchIndex0;
const BYTE* match1 = base + matchIndex1;
U32 offcode;
#if defined(__aarch64__)
PREFETCH_L1(ip0+256);
#endif
hashTable[h0] = current0; /* update hash table */
hashTable[h1] = current1; /* update hash table */
assert(ip0 + 1 == ip1);
if ((offset_1 > 0) & (MEM_read32(repMatch) == MEM_read32(ip2))) {
mLength = (ip2[-1] == repMatch[-1]) ? 1 : 0;
ip0 = ip2 - mLength;
match0 = repMatch - mLength;
mLength += 4;
offcode = 0;
goto _match;
}
if ((matchIndex0 > prefixStartIndex) && MEM_read32(match0) == val0) {
/* found a regular match */
goto _offset;
}
if ((matchIndex1 > prefixStartIndex) && MEM_read32(match1) == val1) {
/* found a regular match after one literal */
ip0 = ip1;
match0 = match1;
goto _offset;
}
{ size_t const step = ((size_t)(ip0-anchor) >> (kSearchStrength - 1)) + stepSize;
assert(step >= 2);
ip0 += step;
ip1 += step;
continue;
}
_offset: /* Requires: ip0, match0 */
/* Compute the offset code */
offset_2 = offset_1;
offset_1 = (U32)(ip0-match0);
offcode = offset_1 + ZSTD_REP_MOVE;
mLength = 4;
/* Count the backwards match length */
while (((ip0>anchor) & (match0>prefixStart))
&& (ip0[-1] == match0[-1])) { ip0--; match0--; mLength++; } /* catch up */
_match: /* Requires: ip0, match0, offcode */
/* Count the forward length */
mLength += ZSTD_count(ip0+mLength, match0+mLength, iend);
ZSTD_storeSeq(seqStore, (size_t)(ip0-anchor), anchor, iend, offcode, mLength-MINMATCH);
/* match found */
ip0 += mLength;
anchor = ip0;
if (ip0 <= ilimit) {
/* Fill Table */
assert(base+current0+2 > istart); /* check base overflow */
hashTable[ZSTD_hashPtr(base+current0+2, hlog, mls)] = current0+2; /* here because current+2 could be > iend-8 */
hashTable[ZSTD_hashPtr(ip0-2, hlog, mls)] = (U32)(ip0-2-base);
if (offset_2 > 0) { /* offset_2==0 means offset_2 is invalidated */
while ( (ip0 <= ilimit) && (MEM_read32(ip0) == MEM_read32(ip0 - offset_2)) ) {
/* store sequence */
size_t const rLength = ZSTD_count(ip0+4, ip0+4-offset_2, iend) + 4;
{ U32 const tmpOff = offset_2; offset_2 = offset_1; offset_1 = tmpOff; } /* swap offset_2 <=> offset_1 */
hashTable[ZSTD_hashPtr(ip0, hlog, mls)] = (U32)(ip0-base);
ip0 += rLength;
ZSTD_storeSeq(seqStore, 0 /*litLen*/, anchor, iend, 0 /*offCode*/, rLength-MINMATCH);
anchor = ip0;
continue; /* faster when present (confirmed on gcc-8) ... (?) */
} } }
ip1 = ip0 + 1;
}
/* save reps for next block */
rep[0] = offset_1 ? offset_1 : offsetSaved;
rep[1] = offset_2 ? offset_2 : offsetSaved;
/* Return the last literals size */
return (size_t)(iend - anchor);
}
size_t ZSTD_compressBlock_fast(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
U32 const mls = ms->cParams.minMatch;
assert(ms->dictMatchState == NULL);
switch(mls)
{
default: /* includes case 3 */
case 4 :
return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 4);
case 5 :
return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 5);
case 6 :
return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 6);
case 7 :
return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 7);
}
}
FORCE_INLINE_TEMPLATE
size_t ZSTD_compressBlock_fast_generic(
size_t ZSTD_compressBlock_fast_dictMatchState_generic(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize,
U32 const mls, ZSTD_dictMode_e const dictMode)
void const* src, size_t srcSize, U32 const mls)
{
const ZSTD_compressionParameters* const cParams = &ms->cParams;
U32* const hashTable = ms->hashTable;
@ -64,46 +225,34 @@ size_t ZSTD_compressBlock_fast_generic(
U32 offsetSaved = 0;
const ZSTD_matchState_t* const dms = ms->dictMatchState;
const ZSTD_compressionParameters* const dictCParams =
dictMode == ZSTD_dictMatchState ?
&dms->cParams : NULL;
const U32* const dictHashTable = dictMode == ZSTD_dictMatchState ?
dms->hashTable : NULL;
const U32 dictStartIndex = dictMode == ZSTD_dictMatchState ?
dms->window.dictLimit : 0;
const BYTE* const dictBase = dictMode == ZSTD_dictMatchState ?
dms->window.base : NULL;
const BYTE* const dictStart = dictMode == ZSTD_dictMatchState ?
dictBase + dictStartIndex : NULL;
const BYTE* const dictEnd = dictMode == ZSTD_dictMatchState ?
dms->window.nextSrc : NULL;
const U32 dictIndexDelta = dictMode == ZSTD_dictMatchState ?
prefixStartIndex - (U32)(dictEnd - dictBase) :
0;
const ZSTD_compressionParameters* const dictCParams = &dms->cParams ;
const U32* const dictHashTable = dms->hashTable;
const U32 dictStartIndex = dms->window.dictLimit;
const BYTE* const dictBase = dms->window.base;
const BYTE* const dictStart = dictBase + dictStartIndex;
const BYTE* const dictEnd = dms->window.nextSrc;
const U32 dictIndexDelta = prefixStartIndex - (U32)(dictEnd - dictBase);
const U32 dictAndPrefixLength = (U32)(ip - prefixStart + dictEnd - dictStart);
const U32 dictHLog = dictMode == ZSTD_dictMatchState ?
dictCParams->hashLog : hlog;
const U32 dictHLog = dictCParams->hashLog;
assert(dictMode == ZSTD_noDict || dictMode == ZSTD_dictMatchState);
/* if a dictionary is still attached, it necessarily means that
* it is within window size. So we just check it. */
const U32 maxDistance = 1U << cParams->windowLog;
const U32 endIndex = (U32)((size_t)(ip - base) + srcSize);
assert(endIndex - prefixStartIndex <= maxDistance);
(void)maxDistance; (void)endIndex; /* these variables are not used when assert() is disabled */
/* otherwise, we would get index underflow when translating a dict index
* into a local index */
assert(dictMode != ZSTD_dictMatchState
|| prefixStartIndex >= (U32)(dictEnd - dictBase));
/* ensure there will be no no underflow
* when translating a dict index into a local index */
assert(prefixStartIndex >= (U32)(dictEnd - dictBase));
/* init */
DEBUGLOG(5, "ZSTD_compressBlock_fast_dictMatchState_generic");
ip += (dictAndPrefixLength == 0);
if (dictMode == ZSTD_noDict) {
U32 const maxRep = (U32)(ip - prefixStart);
if (offset_2 > maxRep) offsetSaved = offset_2, offset_2 = 0;
if (offset_1 > maxRep) offsetSaved = offset_1, offset_1 = 0;
}
if (dictMode == ZSTD_dictMatchState) {
/* dictMatchState repCode checks don't currently handle repCode == 0
* disabling. */
assert(offset_1 <= dictAndPrefixLength);
assert(offset_2 <= dictAndPrefixLength);
}
/* Main Search Loop */
while (ip < ilimit) { /* < instead of <=, because repcode check at (ip+1) */
@ -113,26 +262,18 @@ size_t ZSTD_compressBlock_fast_generic(
U32 const matchIndex = hashTable[h];
const BYTE* match = base + matchIndex;
const U32 repIndex = current + 1 - offset_1;
const BYTE* repMatch = (dictMode == ZSTD_dictMatchState
&& repIndex < prefixStartIndex) ?
const BYTE* repMatch = (repIndex < prefixStartIndex) ?
dictBase + (repIndex - dictIndexDelta) :
base + repIndex;
hashTable[h] = current; /* update hash table */
if ( (dictMode == ZSTD_dictMatchState)
&& ((U32)((prefixStartIndex-1) - repIndex) >= 3) /* intentional underflow : ensure repIndex isn't overlapping dict + prefix */
if ( ((U32)((prefixStartIndex-1) - repIndex) >= 3) /* intentional underflow : ensure repIndex isn't overlapping dict + prefix */
&& (MEM_read32(repMatch) == MEM_read32(ip+1)) ) {
const BYTE* const repMatchEnd = repIndex < prefixStartIndex ? dictEnd : iend;
mLength = ZSTD_count_2segments(ip+1+4, repMatch+4, iend, repMatchEnd, prefixStart) + 4;
ip++;
ZSTD_storeSeq(seqStore, ip-anchor, anchor, 0, mLength-MINMATCH);
} else if ( dictMode == ZSTD_noDict
&& ((offset_1 > 0) & (MEM_read32(ip+1-offset_1) == MEM_read32(ip+1)))) {
mLength = ZSTD_count(ip+1+4, ip+1+4-offset_1, iend) + 4;
ip++;
ZSTD_storeSeq(seqStore, ip-anchor, anchor, 0, mLength-MINMATCH);
ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, 0, mLength-MINMATCH);
} else if ( (matchIndex <= prefixStartIndex) ) {
if (dictMode == ZSTD_dictMatchState) {
size_t const dictHash = ZSTD_hashPtr(ip, dictHLog, mls);
U32 const dictMatchIndex = dictHashTable[dictHash];
const BYTE* dictMatch = dictBase + dictMatchIndex;
@ -151,12 +292,7 @@ size_t ZSTD_compressBlock_fast_generic(
} /* catch up */
offset_2 = offset_1;
offset_1 = offset;
ZSTD_storeSeq(seqStore, ip-anchor, anchor, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
}
} else {
assert(stepSize >= 1);
ip += ((ip-anchor) >> kSearchStrength) + stepSize;
continue;
ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
}
} else if (MEM_read32(match) != MEM_read32(ip)) {
/* it's not a match, and we're not going to check the dictionary */
@ -171,7 +307,7 @@ size_t ZSTD_compressBlock_fast_generic(
&& (ip[-1] == match[-1])) { ip--; match--; mLength++; } /* catch up */
offset_2 = offset_1;
offset_1 = offset;
ZSTD_storeSeq(seqStore, ip-anchor, anchor, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
}
/* match found */
@ -185,7 +321,6 @@ size_t ZSTD_compressBlock_fast_generic(
hashTable[ZSTD_hashPtr(ip-2, hlog, mls)] = (U32)(ip-2-base);
/* check immediate repcode */
if (dictMode == ZSTD_dictMatchState) {
while (ip <= ilimit) {
U32 const current2 = (U32)(ip-base);
U32 const repIndex2 = current2 - offset_2;
@ -197,7 +332,7 @@ size_t ZSTD_compressBlock_fast_generic(
const BYTE* const repEnd2 = repIndex2 < prefixStartIndex ? dictEnd : iend;
size_t const repLength2 = ZSTD_count_2segments(ip+4, repMatch2+4, iend, repEnd2, prefixStart) + 4;
U32 tmpOffset = offset_2; offset_2 = offset_1; offset_1 = tmpOffset; /* swap offset_2 <=> offset_1 */
ZSTD_storeSeq(seqStore, 0, anchor, 0, repLength2-MINMATCH);
ZSTD_storeSeq(seqStore, 0, anchor, iend, 0, repLength2-MINMATCH);
hashTable[ZSTD_hashPtr(ip, hlog, mls)] = current2;
ip += repLength2;
anchor = ip;
@ -206,69 +341,33 @@ size_t ZSTD_compressBlock_fast_generic(
break;
}
}
if (dictMode == ZSTD_noDict) {
while ( (ip <= ilimit)
&& ( (offset_2>0)
& (MEM_read32(ip) == MEM_read32(ip - offset_2)) )) {
/* store sequence */
size_t const rLength = ZSTD_count(ip+4, ip+4-offset_2, iend) + 4;
U32 const tmpOff = offset_2; offset_2 = offset_1; offset_1 = tmpOff; /* swap offset_2 <=> offset_1 */
hashTable[ZSTD_hashPtr(ip, hlog, mls)] = (U32)(ip-base);
ZSTD_storeSeq(seqStore, 0, anchor, 0, rLength-MINMATCH);
ip += rLength;
anchor = ip;
continue; /* faster when present ... (?) */
} } } }
}
/* save reps for next block */
rep[0] = offset_1 ? offset_1 : offsetSaved;
rep[1] = offset_2 ? offset_2 : offsetSaved;
/* Return the last literals size */
return iend - anchor;
}
size_t ZSTD_compressBlock_fast(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
ZSTD_compressionParameters const* cParams = &ms->cParams;
U32 const mls = cParams->searchLength;
assert(ms->dictMatchState == NULL);
switch(mls)
{
default: /* includes case 3 */
case 4 :
return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 4, ZSTD_noDict);
case 5 :
return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 5, ZSTD_noDict);
case 6 :
return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 6, ZSTD_noDict);
case 7 :
return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 7, ZSTD_noDict);
}
return (size_t)(iend - anchor);
}
size_t ZSTD_compressBlock_fast_dictMatchState(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
ZSTD_compressionParameters const* cParams = &ms->cParams;
U32 const mls = cParams->searchLength;
U32 const mls = ms->cParams.minMatch;
assert(ms->dictMatchState != NULL);
switch(mls)
{
default: /* includes case 3 */
case 4 :
return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 4, ZSTD_dictMatchState);
return ZSTD_compressBlock_fast_dictMatchState_generic(ms, seqStore, rep, src, srcSize, 4);
case 5 :
return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 5, ZSTD_dictMatchState);
return ZSTD_compressBlock_fast_dictMatchState_generic(ms, seqStore, rep, src, srcSize, 5);
case 6 :
return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 6, ZSTD_dictMatchState);
return ZSTD_compressBlock_fast_dictMatchState_generic(ms, seqStore, rep, src, srcSize, 6);
case 7 :
return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, 7, ZSTD_dictMatchState);
return ZSTD_compressBlock_fast_dictMatchState_generic(ms, seqStore, rep, src, srcSize, 7);
}
}
@ -287,15 +386,24 @@ static size_t ZSTD_compressBlock_fast_extDict_generic(
const BYTE* const istart = (const BYTE*)src;
const BYTE* ip = istart;
const BYTE* anchor = istart;
const U32 dictStartIndex = ms->window.lowLimit;
const U32 endIndex = (U32)((size_t)(istart - base) + srcSize);
const U32 lowLimit = ZSTD_getLowestMatchIndex(ms, endIndex, cParams->windowLog);
const U32 dictStartIndex = lowLimit;
const BYTE* const dictStart = dictBase + dictStartIndex;
const U32 prefixStartIndex = ms->window.dictLimit;
const U32 dictLimit = ms->window.dictLimit;
const U32 prefixStartIndex = dictLimit < lowLimit ? lowLimit : dictLimit;
const BYTE* const prefixStart = base + prefixStartIndex;
const BYTE* const dictEnd = dictBase + prefixStartIndex;
const BYTE* const iend = istart + srcSize;
const BYTE* const ilimit = iend - 8;
U32 offset_1=rep[0], offset_2=rep[1];
DEBUGLOG(5, "ZSTD_compressBlock_fast_extDict_generic (offset_1=%u)", offset_1);
/* switch to "regular" variant if extDict is invalidated due to maxDistance */
if (prefixStartIndex == dictStartIndex)
return ZSTD_compressBlock_fast_generic(ms, seqStore, rep, src, srcSize, mls);
/* Search Loop */
while (ip < ilimit) { /* < instead of <=, because (ip+1) */
const size_t h = ZSTD_hashPtr(ip, hlog, mls);
@ -306,16 +414,18 @@ static size_t ZSTD_compressBlock_fast_extDict_generic(
const U32 repIndex = current + 1 - offset_1;
const BYTE* const repBase = repIndex < prefixStartIndex ? dictBase : base;
const BYTE* const repMatch = repBase + repIndex;
size_t mLength;
hashTable[h] = current; /* update hash table */
DEBUGLOG(7, "offset_1 = %u , current = %u", offset_1, current);
assert(offset_1 <= current +1); /* check repIndex */
if ( (((U32)((prefixStartIndex-1) - repIndex) >= 3) /* intentional underflow */ & (repIndex > dictStartIndex))
&& (MEM_read32(repMatch) == MEM_read32(ip+1)) ) {
const BYTE* repMatchEnd = repIndex < prefixStartIndex ? dictEnd : iend;
mLength = ZSTD_count_2segments(ip+1+4, repMatch+4, iend, repMatchEnd, prefixStart) + 4;
const BYTE* const repMatchEnd = repIndex < prefixStartIndex ? dictEnd : iend;
size_t const rLength = ZSTD_count_2segments(ip+1 +4, repMatch +4, iend, repMatchEnd, prefixStart) + 4;
ip++;
ZSTD_storeSeq(seqStore, ip-anchor, anchor, 0, mLength-MINMATCH);
ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, 0, rLength-MINMATCH);
ip += rLength;
anchor = ip;
} else {
if ( (matchIndex < dictStartIndex) ||
(MEM_read32(match) != MEM_read32(ip)) ) {
@ -323,20 +433,16 @@ static size_t ZSTD_compressBlock_fast_extDict_generic(
ip += ((ip-anchor) >> kSearchStrength) + stepSize;
continue;
}
{ const BYTE* matchEnd = matchIndex < prefixStartIndex ? dictEnd : iend;
const BYTE* lowMatchPtr = matchIndex < prefixStartIndex ? dictStart : prefixStart;
U32 offset;
mLength = ZSTD_count_2segments(ip+4, match+4, iend, matchEnd, prefixStart) + 4;
{ const BYTE* const matchEnd = matchIndex < prefixStartIndex ? dictEnd : iend;
const BYTE* const lowMatchPtr = matchIndex < prefixStartIndex ? dictStart : prefixStart;
U32 const offset = current - matchIndex;
size_t mLength = ZSTD_count_2segments(ip+4, match+4, iend, matchEnd, prefixStart) + 4;
while (((ip>anchor) & (match>lowMatchPtr)) && (ip[-1] == match[-1])) { ip--; match--; mLength++; } /* catch up */
offset = current - matchIndex;
offset_2 = offset_1;
offset_1 = offset;
ZSTD_storeSeq(seqStore, ip-anchor, anchor, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
} }
/* found a match : store it */
offset_2 = offset_1; offset_1 = offset; /* update offset history */
ZSTD_storeSeq(seqStore, (size_t)(ip-anchor), anchor, iend, offset + ZSTD_REP_MOVE, mLength-MINMATCH);
ip += mLength;
anchor = ip;
} }
if (ip <= ilimit) {
/* Fill Table */
@ -346,13 +452,13 @@ static size_t ZSTD_compressBlock_fast_extDict_generic(
while (ip <= ilimit) {
U32 const current2 = (U32)(ip-base);
U32 const repIndex2 = current2 - offset_2;
const BYTE* repMatch2 = repIndex2 < prefixStartIndex ? dictBase + repIndex2 : base + repIndex2;
const BYTE* const repMatch2 = repIndex2 < prefixStartIndex ? dictBase + repIndex2 : base + repIndex2;
if ( (((U32)((prefixStartIndex-1) - repIndex2) >= 3) & (repIndex2 > dictStartIndex)) /* intentional overflow */
&& (MEM_read32(repMatch2) == MEM_read32(ip)) ) {
const BYTE* const repEnd2 = repIndex2 < prefixStartIndex ? dictEnd : iend;
size_t const repLength2 = ZSTD_count_2segments(ip+4, repMatch2+4, iend, repEnd2, prefixStart) + 4;
U32 tmpOffset = offset_2; offset_2 = offset_1; offset_1 = tmpOffset; /* swap offset_2 <=> offset_1 */
ZSTD_storeSeq(seqStore, 0, anchor, 0, repLength2-MINMATCH);
{ U32 const tmpOffset = offset_2; offset_2 = offset_1; offset_1 = tmpOffset; } /* swap offset_2 <=> offset_1 */
ZSTD_storeSeq(seqStore, 0 /*litlen*/, anchor, iend, 0 /*offcode*/, repLength2-MINMATCH);
hashTable[ZSTD_hashPtr(ip, hlog, mls)] = current2;
ip += repLength2;
anchor = ip;
@ -366,7 +472,7 @@ static size_t ZSTD_compressBlock_fast_extDict_generic(
rep[1] = offset_2;
/* Return the last literals size */
return iend - anchor;
return (size_t)(iend - anchor);
}
@ -374,8 +480,7 @@ size_t ZSTD_compressBlock_fast_extDict(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
ZSTD_compressionParameters const* cParams = &ms->cParams;
U32 const mls = cParams->searchLength;
U32 const mls = ms->cParams.minMatch;
switch(mls)
{
default: /* includes case 3 */

2
drivers/filesystems/btrfs/zstd/zstd_fast.h
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the

232
drivers/filesystems/btrfs/zstd/zstd_internal.h
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -19,6 +19,9 @@
/*-*************************************
* Dependencies
***************************************/
#ifdef __aarch64__
#include <arm_neon.h>
#endif
#include "compiler.h"
#include "mem.h"
#include "debug.h" /* assert, DEBUGLOG, RAWLOG, g_debuglevel */
@ -34,13 +37,15 @@
#endif
#include "../xxhash.h" /* XXH_reset, update, digest */
#if defined (__cplusplus)
extern "C" {
#endif
/* ---- static assert (debug) --- */
#define ZSTD_STATIC_ASSERT(c) DEBUG_STATIC_ASSERT(c)
#define ZSTD_isError ERR_isError /* for inlining */
#define FSE_isError ERR_isError
#define HUF_isError ERR_isError
/*-*************************************
@ -50,8 +55,81 @@ extern "C" {
#undef MAX
#define MIN(a,b) ((a)<(b) ? (a) : (b))
#define MAX(a,b) ((a)>(b) ? (a) : (b))
#define CHECK_F(f) { size_t const errcod = f; if (ERR_isError(errcod)) return errcod; } /* check and Forward error code */
#define CHECK_E(f, e) { size_t const errcod = f; if (ERR_isError(errcod)) return ERROR(e); } /* check and send Error code */
/**
* Ignore: this is an internal helper.
*
* This is a helper function to help force C99-correctness during compilation.
* Under strict compilation modes, variadic macro arguments can't be empty.
* However, variadic function arguments can be. Using a function therefore lets
* us statically check that at least one (string) argument was passed,
* independent of the compilation flags.
*/
static INLINE_KEYWORD UNUSED_ATTR
void _force_has_format_string(const char *format, ...) {
(void)format;
}
/**
* Ignore: this is an internal helper.
*
* We want to force this function invocation to be syntactically correct, but
* we don't want to force runtime evaluation of its arguments.
*/
#define _FORCE_HAS_FORMAT_STRING(...) \
if (0) { \
_force_has_format_string(__VA_ARGS__); \
}
/**
* Return the specified error if the condition evaluates to true.
*
* In debug modes, prints additional information.
* In order to do that (particularly, printing the conditional that failed),
* this can't just wrap RETURN_ERROR().
*/
#define RETURN_ERROR_IF(cond, err, ...) \
if (cond) { \
RAWLOG(3, "%s:%d: ERROR!: check %s failed, returning %s", \
__FILE__, __LINE__, ZSTD_QUOTE(cond), ZSTD_QUOTE(ERROR(err))); \
_FORCE_HAS_FORMAT_STRING(__VA_ARGS__); \
RAWLOG(3, ": " __VA_ARGS__); \
RAWLOG(3, "\n"); \
return ERROR(err); \
}
/**
* Unconditionally return the specified error.
*
* In debug modes, prints additional information.
*/
#define RETURN_ERROR(err, ...) \
do { \
RAWLOG(3, "%s:%d: ERROR!: unconditional check failed, returning %s", \
__FILE__, __LINE__, ZSTD_QUOTE(ERROR(err))); \
_FORCE_HAS_FORMAT_STRING(__VA_ARGS__); \
RAWLOG(3, ": " __VA_ARGS__); \
RAWLOG(3, "\n"); \
return ERROR(err); \
} while(0);
/**
* If the provided expression evaluates to an error code, returns that error code.
*
* In debug modes, prints additional information.
*/
#define FORWARD_IF_ERROR(err, ...) \
do { \
size_t const err_code = (err); \
if (ERR_isError(err_code)) { \
RAWLOG(3, "%s:%d: ERROR!: forwarding error in %s: %s", \
__FILE__, __LINE__, ZSTD_QUOTE(err), ERR_getErrorName(err_code)); \
_FORCE_HAS_FORMAT_STRING(__VA_ARGS__); \
RAWLOG(3, ": " __VA_ARGS__); \
RAWLOG(3, "\n"); \
return err_code; \
} \
} while(0);
/*-*************************************
@ -75,7 +153,6 @@ static const U32 repStartValue[ZSTD_REP_NUM] = { 1, 4, 8 };
#define BIT0 1
#define ZSTD_WINDOWLOG_ABSOLUTEMIN 10
#define ZSTD_WINDOWLOG_DEFAULTMAX 27 /* Default maximum allowed window log */
static const size_t ZSTD_fcs_fieldSize[4] = { 0, 2, 4, 8 };
static const size_t ZSTD_did_fieldSize[4] = { 0, 1, 2, 4 };
@ -85,6 +162,8 @@ static const size_t ZSTD_did_fieldSize[4] = { 0, 1, 2, 4 };
static const size_t ZSTD_blockHeaderSize = ZSTD_BLOCKHEADERSIZE;
typedef enum { bt_raw, bt_rle, bt_compressed, bt_reserved } blockType_e;
#define ZSTD_FRAMECHECKSUMSIZE 4
#define MIN_SEQUENCES_SIZE 1 /* nbSeq==0 */
#define MIN_CBLOCK_SIZE (1 /*litCSize*/ + 1 /* RLE or RAW */ + MIN_SEQUENCES_SIZE /* nbSeq==0 */) /* for a non-null block */
@ -148,32 +227,99 @@ static const U32 OF_defaultNormLog = OF_DEFAULTNORMLOG;
/*-*******************************************
* Shared functions to include for inlining
*********************************************/
static void ZSTD_copy8(void* dst, const void* src) { memcpy(dst, src, 8); }
static void ZSTD_copy8(void* dst, const void* src) {
#ifdef __aarch64__
vst1_u8((uint8_t*)dst, vld1_u8((const uint8_t*)src));
#else
memcpy(dst, src, 8);
#endif
}
#define COPY8(d,s) { ZSTD_copy8(d,s); d+=8; s+=8; }
static void ZSTD_copy16(void* dst, const void* src) {
#ifdef __aarch64__
vst1q_u8((uint8_t*)dst, vld1q_u8((const uint8_t*)src));
#else
memcpy(dst, src, 16);
#endif
}
#define COPY16(d,s) { ZSTD_copy16(d,s); d+=16; s+=16; }
#define WILDCOPY_OVERLENGTH 32
#define WILDCOPY_VECLEN 16
typedef enum {
ZSTD_no_overlap,
ZSTD_overlap_src_before_dst
/* ZSTD_overlap_dst_before_src, */
} ZSTD_overlap_e;
/*! ZSTD_wildcopy() :
* custom version of memcpy(), can overwrite up to WILDCOPY_OVERLENGTH bytes (if length==0) */
#define WILDCOPY_OVERLENGTH 8
MEM_STATIC void ZSTD_wildcopy(void* dst, const void* src, ptrdiff_t length)
* Custom version of memcpy(), can over read/write up to WILDCOPY_OVERLENGTH bytes (if length==0)
* @param ovtype controls the overlap detection
* - ZSTD_no_overlap: The source and destination are guaranteed to be at least WILDCOPY_VECLEN bytes apart.
* - ZSTD_overlap_src_before_dst: The src and dst may overlap, but they MUST be at least 8 bytes apart.
* The src buffer must be before the dst buffer.
*/
MEM_STATIC FORCE_INLINE_ATTR
void ZSTD_wildcopy(void* dst, const void* src, ptrdiff_t length, ZSTD_overlap_e const ovtype)
{
ptrdiff_t diff = (BYTE*)dst - (const BYTE*)src;
const BYTE* ip = (const BYTE*)src;
BYTE* op = (BYTE*)dst;
BYTE* const oend = op + length;
do
assert(diff >= 8 || (ovtype == ZSTD_no_overlap && diff <= -WILDCOPY_VECLEN));
if (ovtype == ZSTD_overlap_src_before_dst && diff < WILDCOPY_VECLEN) {
/* Handle short offset copies. */
do {
COPY8(op, ip)
} while (op < oend);
} else {
assert(diff >= WILDCOPY_VECLEN || diff <= -WILDCOPY_VECLEN);
/* Separate out the first COPY16() call because the copy length is
* almost certain to be short, so the branches have different
* probabilities. Since it is almost certain to be short, only do
* one COPY16() in the first call. Then, do two calls per loop since
* at that point it is more likely to have a high trip count.
*/
#ifndef __aarch64__
do {
COPY16(op, ip);
}
while (op < oend);
#else
COPY16(op, ip);
if (op >= oend) return;
do {
COPY16(op, ip);
COPY16(op, ip);
}
while (op < oend);
#endif
}
}
MEM_STATIC void ZSTD_wildcopy_e(void* dst, const void* src, void* dstEnd) /* should be faster for decoding, but strangely, not verified on all platform */
MEM_STATIC size_t ZSTD_limitCopy(void* dst, size_t dstCapacity, const void* src, size_t srcSize)
{
const BYTE* ip = (const BYTE*)src;
BYTE* op = (BYTE*)dst;
BYTE* const oend = (BYTE*)dstEnd;
do
COPY8(op, ip)
while (op < oend);
size_t const length = MIN(dstCapacity, srcSize);
if (length > 0) {
memcpy(dst, src, length);
}
return length;
}
/* define "workspace is too large" as this number of times larger than needed */
#define ZSTD_WORKSPACETOOLARGE_FACTOR 3
/* when workspace is continuously too large
* during at least this number of times,
* context's memory usage is considered wasteful,
* because it's sized to handle a worst case scenario which rarely happens.
* In which case, resize it down to free some memory */
#define ZSTD_WORKSPACETOOLARGE_MAXDURATION 128
/*-*******************************************
* Private declarations
@ -198,6 +344,42 @@ typedef struct {
U32 longLengthPos;
} seqStore_t;
typedef struct {
U32 litLength;
U32 matchLength;
} ZSTD_sequenceLength;
/**
* Returns the ZSTD_sequenceLength for the given sequences. It handles the decoding of long sequences
* indicated by longLengthPos and longLengthID, and adds MINMATCH back to matchLength.
*/
MEM_STATIC ZSTD_sequenceLength ZSTD_getSequenceLength(seqStore_t const* seqStore, seqDef const* seq)
{
ZSTD_sequenceLength seqLen;
seqLen.litLength = seq->litLength;
seqLen.matchLength = seq->matchLength + MINMATCH;
if (seqStore->longLengthPos == (U32)(seq - seqStore->sequencesStart)) {
if (seqStore->longLengthID == 1) {
seqLen.litLength += 0xFFFF;
}
if (seqStore->longLengthID == 2) {
seqLen.matchLength += 0xFFFF;
}
}
return seqLen;
}
/**
* Contains the compressed frame size and an upper-bound for the decompressed frame size.
* Note: before using `compressedSize`, check for errors using ZSTD_isError().
* similarly, before using `decompressedBound`, check for errors using:
* `decompressedBound != ZSTD_CONTENTSIZE_ERROR`
*/
typedef struct {
size_t compressedSize;
unsigned long long decompressedBound;
} ZSTD_frameSizeInfo; /* decompress & legacy */
const seqStore_t* ZSTD_getSeqStore(const ZSTD_CCtx* ctx); /* compress & dictBuilder */
void ZSTD_seqToCodes(const seqStore_t* seqStorePtr); /* compress, dictBuilder, decodeCorpus (shouldn't get its definition from here) */
@ -213,10 +395,11 @@ MEM_STATIC U32 ZSTD_highbit32(U32 val) /* compress, dictBuilder, decodeCorpus
{
# if defined(_MSC_VER) /* Visual */
unsigned long r=0;
_BitScanReverse(&r, val);
return (unsigned)r;
return _BitScanReverse(&r, val) ? (unsigned)r : 0;
# elif defined(__GNUC__) && (__GNUC__ >= 3) /* GCC Intrinsic */
return 31 - __builtin_clz(val);
return __builtin_clz (val) ^ 31;
# elif defined(__ICCARM__) /* IAR Intrinsic */
return 31 - __CLZ(val);
# else /* Software version */
static const U32 DeBruijnClz[32] = { 0, 9, 1, 10, 13, 21, 2, 29, 11, 14, 16, 18, 22, 25, 3, 30, 8, 12, 20, 28, 15, 17, 24, 7, 19, 27, 23, 6, 26, 5, 4, 31 };
U32 v = val;
@ -242,7 +425,7 @@ typedef struct {
blockType_e blockType;
U32 lastBlock;
U32 origSize;
} blockProperties_t;
} blockProperties_t; /* declared here for decompress and fullbench */
/*! ZSTD_getcBlockSize() :
* Provides the size of compressed block from block header `src` */
@ -250,6 +433,13 @@ typedef struct {
size_t ZSTD_getcBlockSize(const void* src, size_t srcSize,
blockProperties_t* bpPtr);
/*! ZSTD_decodeSeqHeaders() :
* decode sequence header from src */
/* Used by: decompress, fullbench (does not get its definition from here) */
size_t ZSTD_decodeSeqHeaders(ZSTD_DCtx* dctx, int* nbSeqPtr,
const void* src, size_t srcSize);
#if defined (__cplusplus)
}
#endif

153
drivers/filesystems/btrfs/zstd/zstd_lazy.c
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -63,7 +63,8 @@ ZSTD_updateDUBT(ZSTD_matchState_t* ms,
static void
ZSTD_insertDUBT1(ZSTD_matchState_t* ms,
U32 current, const BYTE* inputEnd,
U32 nbCompares, U32 btLow, const ZSTD_dictMode_e dictMode)
U32 nbCompares, U32 btLow,
const ZSTD_dictMode_e dictMode)
{
const ZSTD_compressionParameters* const cParams = &ms->cParams;
U32* const bt = ms->chainTable;
@ -80,9 +81,12 @@ ZSTD_insertDUBT1(ZSTD_matchState_t* ms,
const BYTE* match;
U32* smallerPtr = bt + 2*(current&btMask);
U32* largerPtr = smallerPtr + 1;
U32 matchIndex = *smallerPtr;
U32 matchIndex = *smallerPtr; /* this candidate is unsorted : next sorted candidate is reached through *smallerPtr, while *largerPtr contains previous unsorted candidate (which is already saved and can be overwritten) */
U32 dummy32; /* to be nullified at the end */
U32 const windowLow = ms->window.lowLimit;
U32 const windowValid = ms->window.lowLimit;
U32 const maxDistance = 1U << cParams->windowLog;
U32 const windowLow = (current - windowValid > maxDistance) ? current - maxDistance : windowValid;
DEBUGLOG(8, "ZSTD_insertDUBT1(%u) (dictLimit=%u, lowLimit=%u)",
current, dictLimit, windowLow);
@ -93,6 +97,9 @@ ZSTD_insertDUBT1(ZSTD_matchState_t* ms,
U32* const nextPtr = bt + 2*(matchIndex & btMask);
size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */
assert(matchIndex < current);
/* note : all candidates are now supposed sorted,
* but it's still possible to have nextPtr[1] == ZSTD_DUBT_UNSORTED_MARK
* when a real index has the same value as ZSTD_DUBT_UNSORTED_MARK */
if ( (dictMode != ZSTD_extDict)
|| (matchIndex+matchLength >= dictLimit) /* both in current segment*/
@ -108,7 +115,7 @@ ZSTD_insertDUBT1(ZSTD_matchState_t* ms,
match = dictBase + matchIndex;
matchLength += ZSTD_count_2segments(ip+matchLength, match+matchLength, iend, dictEnd, prefixStart);
if (matchIndex+matchLength >= dictLimit)
match = base + matchIndex; /* to prepare for next usage of match[matchLength] */
match = base + matchIndex; /* preparation for next read of match[matchLength] */
}
DEBUGLOG(8, "ZSTD_insertDUBT1: comparing %u with %u : found %u common bytes ",
@ -235,7 +242,7 @@ ZSTD_DUBT_findBestMatch(ZSTD_matchState_t* ms,
const BYTE* const base = ms->window.base;
U32 const current = (U32)(ip-base);
U32 const windowLow = ms->window.lowLimit;
U32 const windowLow = ZSTD_getLowestMatchIndex(ms, current, cParams->windowLog);
U32* const bt = ms->chainTable;
U32 const btLog = cParams->chainLog - 1;
@ -258,7 +265,7 @@ ZSTD_DUBT_findBestMatch(ZSTD_matchState_t* ms,
&& (nbCandidates > 1) ) {
DEBUGLOG(8, "ZSTD_DUBT_findBestMatch: candidate %u is unsorted",
matchIndex);
*unsortedMark = previousCandidate;
*unsortedMark = previousCandidate; /* the unsortedMark becomes a reversed chain, to move up back to original position */
previousCandidate = matchIndex;
matchIndex = *nextCandidate;
nextCandidate = bt + 2*(matchIndex&btMask);
@ -266,11 +273,13 @@ ZSTD_DUBT_findBestMatch(ZSTD_matchState_t* ms,
nbCandidates --;
}
/* nullify last candidate if it's still unsorted
* simplification, detrimental to compression ratio, beneficial for speed */
if ( (matchIndex > unsortLimit)
&& (*unsortedMark==ZSTD_DUBT_UNSORTED_MARK) ) {
DEBUGLOG(7, "ZSTD_DUBT_findBestMatch: nullify last unsorted candidate %u",
matchIndex);
*nextCandidate = *unsortedMark = 0; /* nullify next candidate if it's still unsorted (note : simplification, detrimental to compression ratio, beneficial for speed) */
*nextCandidate = *unsortedMark = 0;
}
/* batch sort stacked candidates */
@ -285,14 +294,14 @@ ZSTD_DUBT_findBestMatch(ZSTD_matchState_t* ms,
}
/* find longest match */
{ size_t commonLengthSmaller=0, commonLengthLarger=0;
{ size_t commonLengthSmaller = 0, commonLengthLarger = 0;
const BYTE* const dictBase = ms->window.dictBase;
const U32 dictLimit = ms->window.dictLimit;
const BYTE* const dictEnd = dictBase + dictLimit;
const BYTE* const prefixStart = base + dictLimit;
U32* smallerPtr = bt + 2*(current&btMask);
U32* largerPtr = bt + 2*(current&btMask) + 1;
U32 matchEndIdx = current+8+1;
U32 matchEndIdx = current + 8 + 1;
U32 dummy32; /* to be nullified at the end */
size_t bestLength = 0;
@ -386,7 +395,7 @@ ZSTD_BtFindBestMatch_selectMLS ( ZSTD_matchState_t* ms,
const BYTE* ip, const BYTE* const iLimit,
size_t* offsetPtr)
{
switch(ms->cParams.searchLength)
switch(ms->cParams.minMatch)
{
default : /* includes case 3 */
case 4 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 4, ZSTD_noDict);
@ -402,7 +411,7 @@ static size_t ZSTD_BtFindBestMatch_dictMatchState_selectMLS (
const BYTE* ip, const BYTE* const iLimit,
size_t* offsetPtr)
{
switch(ms->cParams.searchLength)
switch(ms->cParams.minMatch)
{
default : /* includes case 3 */
case 4 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 4, ZSTD_dictMatchState);
@ -418,7 +427,7 @@ static size_t ZSTD_BtFindBestMatch_extDict_selectMLS (
const BYTE* ip, const BYTE* const iLimit,
size_t* offsetPtr)
{
switch(ms->cParams.searchLength)
switch(ms->cParams.minMatch)
{
default : /* includes case 3 */
case 4 : return ZSTD_BtFindBestMatch(ms, ip, iLimit, offsetPtr, 4, ZSTD_extDict);
@ -433,7 +442,7 @@ static size_t ZSTD_BtFindBestMatch_extDict_selectMLS (
/* *********************************
* Hash Chain
***********************************/
#define NEXT_IN_CHAIN(d, mask) chainTable[(d) & mask]
#define NEXT_IN_CHAIN(d, mask) chainTable[(d) & (mask)]
/* Update chains up to ip (excluded)
Assumption : always within prefix (i.e. not within extDict) */
@ -463,7 +472,7 @@ static U32 ZSTD_insertAndFindFirstIndex_internal(
U32 ZSTD_insertAndFindFirstIndex(ZSTD_matchState_t* ms, const BYTE* ip) {
const ZSTD_compressionParameters* const cParams = &ms->cParams;
return ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, ms->cParams.searchLength);
return ZSTD_insertAndFindFirstIndex_internal(ms, cParams, ip, ms->cParams.minMatch);
}
@ -484,8 +493,12 @@ size_t ZSTD_HcFindBestMatch_generic (
const U32 dictLimit = ms->window.dictLimit;
const BYTE* const prefixStart = base + dictLimit;
const BYTE* const dictEnd = dictBase + dictLimit;
const U32 lowLimit = ms->window.lowLimit;
const U32 current = (U32)(ip-base);
const U32 maxDistance = 1U << cParams->windowLog;
const U32 lowestValid = ms->window.lowLimit;
const U32 withinMaxDistance = (current - lowestValid > maxDistance) ? current - maxDistance : lowestValid;
const U32 isDictionary = (ms->loadedDictEnd != 0);
const U32 lowLimit = isDictionary ? lowestValid : withinMaxDistance;
const U32 minChain = current > chainSize ? current - chainSize : 0;
U32 nbAttempts = 1U << cParams->searchLog;
size_t ml=4-1;
@ -497,6 +510,7 @@ size_t ZSTD_HcFindBestMatch_generic (
size_t currentMl=0;
if ((dictMode != ZSTD_extDict) || matchIndex >= dictLimit) {
const BYTE* const match = base + matchIndex;
assert(matchIndex >= dictLimit); /* ensures this is true if dictMode != ZSTD_extDict */
if (match[ml] == ip[ml]) /* potentially better */
currentMl = ZSTD_count(ip, match, iLimit);
} else {
@ -559,7 +573,7 @@ FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_selectMLS (
const BYTE* ip, const BYTE* const iLimit,
size_t* offsetPtr)
{
switch(ms->cParams.searchLength)
switch(ms->cParams.minMatch)
{
default : /* includes case 3 */
case 4 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 4, ZSTD_noDict);
@ -575,7 +589,7 @@ static size_t ZSTD_HcFindBestMatch_dictMatchState_selectMLS (
const BYTE* ip, const BYTE* const iLimit,
size_t* offsetPtr)
{
switch(ms->cParams.searchLength)
switch(ms->cParams.minMatch)
{
default : /* includes case 3 */
case 4 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 4, ZSTD_dictMatchState);
@ -591,7 +605,7 @@ FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_extDict_selectMLS (
const BYTE* ip, const BYTE* const iLimit,
size_t* offsetPtr)
{
switch(ms->cParams.searchLength)
switch(ms->cParams.minMatch)
{
default : /* includes case 3 */
case 4 : return ZSTD_HcFindBestMatch_generic(ms, ip, iLimit, offsetPtr, 4, ZSTD_extDict);
@ -605,12 +619,14 @@ FORCE_INLINE_TEMPLATE size_t ZSTD_HcFindBestMatch_extDict_selectMLS (
/* *******************************
* Common parser - lazy strategy
*********************************/
FORCE_INLINE_TEMPLATE
size_t ZSTD_compressBlock_lazy_generic(
typedef enum { search_hashChain, search_binaryTree } searchMethod_e;
FORCE_INLINE_TEMPLATE size_t
ZSTD_compressBlock_lazy_generic(
ZSTD_matchState_t* ms, seqStore_t* seqStore,
U32 rep[ZSTD_REP_NUM],
const void* src, size_t srcSize,
const U32 searchMethod, const U32 depth,
const searchMethod_e searchMethod, const U32 depth,
ZSTD_dictMode_e const dictMode)
{
const BYTE* const istart = (const BYTE*)src;
@ -626,8 +642,10 @@ size_t ZSTD_compressBlock_lazy_generic(
ZSTD_matchState_t* ms,
const BYTE* ip, const BYTE* iLimit, size_t* offsetPtr);
searchMax_f const searchMax = dictMode == ZSTD_dictMatchState ?
(searchMethod ? ZSTD_BtFindBestMatch_dictMatchState_selectMLS : ZSTD_HcFindBestMatch_dictMatchState_selectMLS) :
(searchMethod ? ZSTD_BtFindBestMatch_selectMLS : ZSTD_HcFindBestMatch_selectMLS);
(searchMethod==search_binaryTree ? ZSTD_BtFindBestMatch_dictMatchState_selectMLS
: ZSTD_HcFindBestMatch_dictMatchState_selectMLS) :
(searchMethod==search_binaryTree ? ZSTD_BtFindBestMatch_selectMLS
: ZSTD_HcFindBestMatch_selectMLS);
U32 offset_1 = rep[0], offset_2 = rep[1], savedOffset=0;
const ZSTD_matchState_t* const dms = ms->dictMatchState;
@ -642,13 +660,16 @@ size_t ZSTD_compressBlock_lazy_generic(
const U32 dictIndexDelta = dictMode == ZSTD_dictMatchState ?
prefixLowestIndex - (U32)(dictEnd - dictBase) :
0;
const U32 dictAndPrefixLength = (U32)(ip - prefixLowest + dictEnd - dictLowest);
const U32 dictAndPrefixLength = (U32)((ip - prefixLowest) + (dictEnd - dictLowest));
DEBUGLOG(5, "ZSTD_compressBlock_lazy_generic (dictMode=%u)", (U32)dictMode);
/* init */
ip += (dictAndPrefixLength == 0);
ms->nextToUpdate3 = ms->nextToUpdate;
if (dictMode == ZSTD_noDict) {
U32 const maxRep = (U32)(ip - prefixLowest);
U32 const current = (U32)(ip - base);
U32 const windowLow = ZSTD_getLowestPrefixIndex(ms, current, ms->cParams.windowLog);
U32 const maxRep = current - windowLow;
if (offset_2 > maxRep) savedOffset = offset_2, offset_2 = 0;
if (offset_1 > maxRep) savedOffset = offset_1, offset_1 = 0;
}
@ -660,6 +681,12 @@ size_t ZSTD_compressBlock_lazy_generic(
}
/* Match Loop */
#if defined(__GNUC__) && defined(__x86_64__)
/* I've measured random a 5% speed loss on levels 5 & 6 (greedy) when the
* code alignment is perturbed. To fix the instability align the loop on 32-bytes.
*/
__asm__(".p2align 5");
#endif
while (ip < ilimit) {
size_t matchLength=0;
size_t offset=0;
@ -793,7 +820,7 @@ size_t ZSTD_compressBlock_lazy_generic(
/* store sequence */
_storeSequence:
{ size_t const litLength = start - anchor;
ZSTD_storeSeq(seqStore, litLength, anchor, (U32)offset, matchLength-MINMATCH);
ZSTD_storeSeq(seqStore, litLength, anchor, iend, (U32)offset, matchLength-MINMATCH);
anchor = ip = start + matchLength;
}
@ -811,7 +838,7 @@ _storeSequence:
const BYTE* const repEnd2 = repIndex < prefixLowestIndex ? dictEnd : iend;
matchLength = ZSTD_count_2segments(ip+4, repMatch+4, iend, repEnd2, prefixLowest) + 4;
offset = offset_2; offset_2 = offset_1; offset_1 = (U32)offset; /* swap offset_2 <=> offset_1 */
ZSTD_storeSeq(seqStore, 0, anchor, 0, matchLength-MINMATCH);
ZSTD_storeSeq(seqStore, 0, anchor, iend, 0, matchLength-MINMATCH);
ip += matchLength;
anchor = ip;
continue;
@ -826,7 +853,7 @@ _storeSequence:
/* store sequence */
matchLength = ZSTD_count(ip+4, ip+4-offset_2, iend) + 4;
offset = offset_2; offset_2 = offset_1; offset_1 = (U32)offset; /* swap repcodes */
ZSTD_storeSeq(seqStore, 0, anchor, 0, matchLength-MINMATCH);
ZSTD_storeSeq(seqStore, 0, anchor, iend, 0, matchLength-MINMATCH);
ip += matchLength;
anchor = ip;
continue; /* faster when present ... (?) */
@ -837,7 +864,7 @@ _storeSequence:
rep[1] = offset_2 ? offset_2 : savedOffset;
/* Return the last literals size */
return iend - anchor;
return (size_t)(iend - anchor);
}
@ -845,56 +872,56 @@ size_t ZSTD_compressBlock_btlazy2(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 1, 2, ZSTD_noDict);
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_binaryTree, 2, ZSTD_noDict);
}
size_t ZSTD_compressBlock_lazy2(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 0, 2, ZSTD_noDict);
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2, ZSTD_noDict);
}
size_t ZSTD_compressBlock_lazy(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 0, 1, ZSTD_noDict);
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1, ZSTD_noDict);
}
size_t ZSTD_compressBlock_greedy(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 0, 0, ZSTD_noDict);
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 0, ZSTD_noDict);
}
size_t ZSTD_compressBlock_btlazy2_dictMatchState(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 1, 2, ZSTD_dictMatchState);
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_binaryTree, 2, ZSTD_dictMatchState);
}
size_t ZSTD_compressBlock_lazy2_dictMatchState(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 0, 2, ZSTD_dictMatchState);
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2, ZSTD_dictMatchState);
}
size_t ZSTD_compressBlock_lazy_dictMatchState(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 0, 1, ZSTD_dictMatchState);
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1, ZSTD_dictMatchState);
}
size_t ZSTD_compressBlock_greedy_dictMatchState(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, 0, 0, ZSTD_dictMatchState);
return ZSTD_compressBlock_lazy_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 0, ZSTD_dictMatchState);
}
@ -903,7 +930,7 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(
ZSTD_matchState_t* ms, seqStore_t* seqStore,
U32 rep[ZSTD_REP_NUM],
const void* src, size_t srcSize,
const U32 searchMethod, const U32 depth)
const searchMethod_e searchMethod, const U32 depth)
{
const BYTE* const istart = (const BYTE*)src;
const BYTE* ip = istart;
@ -912,24 +939,31 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(
const BYTE* const ilimit = iend - 8;
const BYTE* const base = ms->window.base;
const U32 dictLimit = ms->window.dictLimit;
const U32 lowestIndex = ms->window.lowLimit;
const BYTE* const prefixStart = base + dictLimit;
const BYTE* const dictBase = ms->window.dictBase;
const BYTE* const dictEnd = dictBase + dictLimit;
const BYTE* const dictStart = dictBase + lowestIndex;
const BYTE* const dictStart = dictBase + ms->window.lowLimit;
const U32 windowLog = ms->cParams.windowLog;
typedef size_t (*searchMax_f)(
ZSTD_matchState_t* ms,
const BYTE* ip, const BYTE* iLimit, size_t* offsetPtr);
searchMax_f searchMax = searchMethod ? ZSTD_BtFindBestMatch_extDict_selectMLS : ZSTD_HcFindBestMatch_extDict_selectMLS;
searchMax_f searchMax = searchMethod==search_binaryTree ? ZSTD_BtFindBestMatch_extDict_selectMLS : ZSTD_HcFindBestMatch_extDict_selectMLS;
U32 offset_1 = rep[0], offset_2 = rep[1];
DEBUGLOG(5, "ZSTD_compressBlock_lazy_extDict_generic");
/* init */
ms->nextToUpdate3 = ms->nextToUpdate;
ip += (ip == prefixStart);
/* Match Loop */
#if defined(__GNUC__) && defined(__x86_64__)
/* I've measured random a 5% speed loss on levels 5 & 6 (greedy) when the
* code alignment is perturbed. To fix the instability align the loop on 32-bytes.
*/
__asm__(".p2align 5");
#endif
while (ip < ilimit) {
size_t matchLength=0;
size_t offset=0;
@ -937,10 +971,11 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(
U32 current = (U32)(ip-base);
/* check repCode */
{ const U32 repIndex = (U32)(current+1 - offset_1);
{ const U32 windowLow = ZSTD_getLowestMatchIndex(ms, current+1, windowLog);
const U32 repIndex = (U32)(current+1 - offset_1);
const BYTE* const repBase = repIndex < dictLimit ? dictBase : base;
const BYTE* const repMatch = repBase + repIndex;
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > lowestIndex)) /* intentional overflow */
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > windowLow)) /* intentional overflow */
if (MEM_read32(ip+1) == MEM_read32(repMatch)) {
/* repcode detected we should take it */
const BYTE* const repEnd = repIndex < dictLimit ? dictEnd : iend;
@ -967,10 +1002,11 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(
current++;
/* check repCode */
if (offset) {
const U32 windowLow = ZSTD_getLowestMatchIndex(ms, current, windowLog);
const U32 repIndex = (U32)(current - offset_1);
const BYTE* const repBase = repIndex < dictLimit ? dictBase : base;
const BYTE* const repMatch = repBase + repIndex;
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > lowestIndex)) /* intentional overflow */
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > windowLow)) /* intentional overflow */
if (MEM_read32(ip) == MEM_read32(repMatch)) {
/* repcode detected */
const BYTE* const repEnd = repIndex < dictLimit ? dictEnd : iend;
@ -997,10 +1033,11 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(
current++;
/* check repCode */
if (offset) {
const U32 windowLow = ZSTD_getLowestMatchIndex(ms, current, windowLog);
const U32 repIndex = (U32)(current - offset_1);
const BYTE* const repBase = repIndex < dictLimit ? dictBase : base;
const BYTE* const repMatch = repBase + repIndex;
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > lowestIndex)) /* intentional overflow */
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > windowLow)) /* intentional overflow */
if (MEM_read32(ip) == MEM_read32(repMatch)) {
/* repcode detected */
const BYTE* const repEnd = repIndex < dictLimit ? dictEnd : iend;
@ -1035,22 +1072,24 @@ size_t ZSTD_compressBlock_lazy_extDict_generic(
/* store sequence */
_storeSequence:
{ size_t const litLength = start - anchor;
ZSTD_storeSeq(seqStore, litLength, anchor, (U32)offset, matchLength-MINMATCH);
ZSTD_storeSeq(seqStore, litLength, anchor, iend, (U32)offset, matchLength-MINMATCH);
anchor = ip = start + matchLength;
}
/* check immediate repcode */
while (ip <= ilimit) {
const U32 repIndex = (U32)((ip-base) - offset_2);
const U32 repCurrent = (U32)(ip-base);
const U32 windowLow = ZSTD_getLowestMatchIndex(ms, repCurrent, windowLog);
const U32 repIndex = repCurrent - offset_2;
const BYTE* const repBase = repIndex < dictLimit ? dictBase : base;
const BYTE* const repMatch = repBase + repIndex;
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > lowestIndex)) /* intentional overflow */
if (((U32)((dictLimit-1) - repIndex) >= 3) & (repIndex > windowLow)) /* intentional overflow */
if (MEM_read32(ip) == MEM_read32(repMatch)) {
/* repcode detected we should take it */
const BYTE* const repEnd = repIndex < dictLimit ? dictEnd : iend;
matchLength = ZSTD_count_2segments(ip+4, repMatch+4, iend, repEnd, prefixStart) + 4;
offset = offset_2; offset_2 = offset_1; offset_1 = (U32)offset; /* swap offset history */
ZSTD_storeSeq(seqStore, 0, anchor, 0, matchLength-MINMATCH);
ZSTD_storeSeq(seqStore, 0, anchor, iend, 0, matchLength-MINMATCH);
ip += matchLength;
anchor = ip;
continue; /* faster when present ... (?) */
@ -1063,7 +1102,7 @@ _storeSequence:
rep[1] = offset_2;
/* Return the last literals size */
return iend - anchor;
return (size_t)(iend - anchor);
}
@ -1071,7 +1110,7 @@ size_t ZSTD_compressBlock_greedy_extDict(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, 0, 0);
return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 0);
}
size_t ZSTD_compressBlock_lazy_extDict(
@ -1079,7 +1118,7 @@ size_t ZSTD_compressBlock_lazy_extDict(
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, 0, 1);
return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 1);
}
size_t ZSTD_compressBlock_lazy2_extDict(
@ -1087,7 +1126,7 @@ size_t ZSTD_compressBlock_lazy2_extDict(
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, 0, 2);
return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_hashChain, 2);
}
size_t ZSTD_compressBlock_btlazy2_extDict(
@ -1095,5 +1134,5 @@ size_t ZSTD_compressBlock_btlazy2_extDict(
void const* src, size_t srcSize)
{
return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, 1, 2);
return ZSTD_compressBlock_lazy_extDict_generic(ms, seqStore, rep, src, srcSize, search_binaryTree, 2);
}

4
drivers/filesystems/btrfs/zstd/zstd_lazy.h
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -19,7 +19,7 @@ extern "C" {
U32 ZSTD_insertAndFindFirstIndex(ZSTD_matchState_t* ms, const BYTE* ip);
void ZSTD_preserveUnsortedMark (U32* const table, U32 const size, U32 const reducerValue); /*! used in ZSTD_reduceIndex(). pre-emptively increase value of ZSTD_DUBT_UNSORTED_MARK */
void ZSTD_preserveUnsortedMark (U32* const table, U32 const size, U32 const reducerValue); /*! used in ZSTD_reduceIndex(). preemptively increase value of ZSTD_DUBT_UNSORTED_MARK */
size_t ZSTD_compressBlock_btlazy2(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],

119
drivers/filesystems/btrfs/zstd/zstd_ldm.c
View file

@ -1,10 +1,11 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
#include "zstd_ldm.h"
@ -37,8 +38,8 @@ void ZSTD_ldm_adjustParameters(ldmParams_t* params,
params->hashLog = MAX(ZSTD_HASHLOG_MIN, params->windowLog - LDM_HASH_RLOG);
assert(params->hashLog <= ZSTD_HASHLOG_MAX);
}
if (params->hashEveryLog == 0) {
params->hashEveryLog = params->windowLog < params->hashLog
if (params->hashRateLog == 0) {
params->hashRateLog = params->windowLog < params->hashLog
? 0
: params->windowLog - params->hashLog;
}
@ -49,9 +50,9 @@ size_t ZSTD_ldm_getTableSize(ldmParams_t params)
{
size_t const ldmHSize = ((size_t)1) << params.hashLog;
size_t const ldmBucketSizeLog = MIN(params.bucketSizeLog, params.hashLog);
size_t const ldmBucketSize =
((size_t)1) << (params.hashLog - ldmBucketSizeLog);
size_t const totalSize = ldmBucketSize + ldmHSize * sizeof(ldmEntry_t);
size_t const ldmBucketSize = ((size_t)1) << (params.hashLog - ldmBucketSizeLog);
size_t const totalSize = ZSTD_cwksp_alloc_size(ldmBucketSize)
+ ZSTD_cwksp_alloc_size(ldmHSize * sizeof(ldmEntry_t));
return params.enableLdm ? totalSize : 0;
}
@ -119,20 +120,20 @@ static void ZSTD_ldm_insertEntry(ldmState_t* ldmState,
*
* Gets the small hash, checksum, and tag from the rollingHash.
*
* If the tag matches (1 << ldmParams.hashEveryLog)-1, then
* If the tag matches (1 << ldmParams.hashRateLog)-1, then
* creates an ldmEntry from the offset, and inserts it into the hash table.
*
* hBits is the length of the small hash, which is the most significant hBits
* of rollingHash. The checksum is the next 32 most significant bits, followed
* by ldmParams.hashEveryLog bits that make up the tag. */
* by ldmParams.hashRateLog bits that make up the tag. */
static void ZSTD_ldm_makeEntryAndInsertByTag(ldmState_t* ldmState,
U64 const rollingHash,
U32 const hBits,
U32 const offset,
ldmParams_t const ldmParams)
{
U32 const tag = ZSTD_ldm_getTag(rollingHash, hBits, ldmParams.hashEveryLog);
U32 const tagMask = ((U32)1 << ldmParams.hashEveryLog) - 1;
U32 const tag = ZSTD_ldm_getTag(rollingHash, hBits, ldmParams.hashRateLog);
U32 const tagMask = ((U32)1 << ldmParams.hashRateLog) - 1;
if (tag == tagMask) {
U32 const hash = ZSTD_ldm_getSmallHash(rollingHash, hBits);
U32 const checksum = ZSTD_ldm_getChecksum(rollingHash, hBits);
@ -143,56 +144,6 @@ static void ZSTD_ldm_makeEntryAndInsertByTag(ldmState_t* ldmState,
}
}
/** ZSTD_ldm_getRollingHash() :
* Get a 64-bit hash using the first len bytes from buf.
*
* Giving bytes s = s_1, s_2, ... s_k, the hash is defined to be
* H(s) = s_1*(a^(k-1)) + s_2*(a^(k-2)) + ... + s_k*(a^0)
*
* where the constant a is defined to be prime8bytes.
*
* The implementation adds an offset to each byte, so
* H(s) = (s_1 + HASH_CHAR_OFFSET)*(a^(k-1)) + ... */
static U64 ZSTD_ldm_getRollingHash(const BYTE* buf, U32 len)
{
U64 ret = 0;
U32 i;
for (i = 0; i < len; i++) {
ret *= prime8bytes;
ret += buf[i] + LDM_HASH_CHAR_OFFSET;
}
return ret;
}
/** ZSTD_ldm_ipow() :
* Return base^exp. */
static U64 ZSTD_ldm_ipow(U64 base, U64 exp)
{
U64 ret = 1;
while (exp) {
if (exp & 1) { ret *= base; }
exp >>= 1;
base *= base;
}
return ret;
}
U64 ZSTD_ldm_getHashPower(U32 minMatchLength) {
DEBUGLOG(4, "ZSTD_ldm_getHashPower: mml=%u", minMatchLength);
assert(minMatchLength >= ZSTD_LDM_MINMATCH_MIN);
return ZSTD_ldm_ipow(prime8bytes, minMatchLength - 1);
}
/** ZSTD_ldm_updateHash() :
* Updates hash by removing toRemove and adding toAdd. */
static U64 ZSTD_ldm_updateHash(U64 hash, BYTE toRemove, BYTE toAdd, U64 hashPower)
{
hash -= ((toRemove + LDM_HASH_CHAR_OFFSET) * hashPower);
hash *= prime8bytes;
hash += toAdd + LDM_HASH_CHAR_OFFSET;
return hash;
}
/** ZSTD_ldm_countBackwardsMatch() :
* Returns the number of bytes that match backwards before pIn and pMatch.
*
@ -238,6 +189,7 @@ static size_t ZSTD_ldm_fillFastTables(ZSTD_matchState_t* ms,
case ZSTD_btlazy2:
case ZSTD_btopt:
case ZSTD_btultra:
case ZSTD_btultra2:
break;
default:
assert(0); /* not possible : not a valid strategy id */
@ -261,7 +213,7 @@ static U64 ZSTD_ldm_fillLdmHashTable(ldmState_t* state,
const BYTE* cur = lastHashed + 1;
while (cur < iend) {
rollingHash = ZSTD_ldm_updateHash(rollingHash, cur[-1],
rollingHash = ZSTD_rollingHash_rotate(rollingHash, cur[-1],
cur[ldmParams.minMatchLength-1],
state->hashPower);
ZSTD_ldm_makeEntryAndInsertByTag(state,
@ -272,6 +224,20 @@ static U64 ZSTD_ldm_fillLdmHashTable(ldmState_t* state,
return rollingHash;
}
void ZSTD_ldm_fillHashTable(
ldmState_t* state, const BYTE* ip,
const BYTE* iend, ldmParams_t const* params)
{
DEBUGLOG(5, "ZSTD_ldm_fillHashTable");
if ((size_t)(iend - ip) >= params->minMatchLength) {
U64 startingHash = ZSTD_rollingHash_compute(ip, params->minMatchLength);
ZSTD_ldm_fillLdmHashTable(
state, startingHash, ip, iend - params->minMatchLength, state->window.base,
params->hashLog - params->bucketSizeLog,
*params);
}
}
/** ZSTD_ldm_limitTableUpdate() :
*
@ -297,8 +263,8 @@ static size_t ZSTD_ldm_generateSequences_internal(
U64 const hashPower = ldmState->hashPower;
U32 const hBits = params->hashLog - params->bucketSizeLog;
U32 const ldmBucketSize = 1U << params->bucketSizeLog;
U32 const hashEveryLog = params->hashEveryLog;
U32 const ldmTagMask = (1U << params->hashEveryLog) - 1;
U32 const hashRateLog = params->hashRateLog;
U32 const ldmTagMask = (1U << params->hashRateLog) - 1;
/* Prefix and extDict parameters */
U32 const dictLimit = ldmState->window.dictLimit;
U32 const lowestIndex = extDict ? ldmState->window.lowLimit : dictLimit;
@ -324,16 +290,16 @@ static size_t ZSTD_ldm_generateSequences_internal(
size_t forwardMatchLength = 0, backwardMatchLength = 0;
ldmEntry_t* bestEntry = NULL;
if (ip != istart) {
rollingHash = ZSTD_ldm_updateHash(rollingHash, lastHashed[0],
rollingHash = ZSTD_rollingHash_rotate(rollingHash, lastHashed[0],
lastHashed[minMatchLength],
hashPower);
} else {
rollingHash = ZSTD_ldm_getRollingHash(ip, minMatchLength);
rollingHash = ZSTD_rollingHash_compute(ip, minMatchLength);
}
lastHashed = ip;
/* Do not insert and do not look for a match */
if (ZSTD_ldm_getTag(rollingHash, hBits, hashEveryLog) != ldmTagMask) {
if (ZSTD_ldm_getTag(rollingHash, hBits, hashRateLog) != ldmTagMask) {
ip++;
continue;
}
@ -478,7 +444,7 @@ size_t ZSTD_ldm_generateSequences(
*/
assert(ldmState->window.nextSrc >= (BYTE const*)src + srcSize);
/* The input could be very large (in zstdmt), so it must be broken up into
* chunks to enforce the maximmum distance and handle overflow correction.
* chunks to enforce the maximum distance and handle overflow correction.
*/
assert(sequences->pos <= sequences->size);
assert(sequences->size <= sequences->capacity);
@ -496,8 +462,10 @@ size_t ZSTD_ldm_generateSequences(
if (ZSTD_window_needOverflowCorrection(ldmState->window, chunkEnd)) {
U32 const ldmHSize = 1U << params->hashLog;
U32 const correction = ZSTD_window_correctOverflow(
&ldmState->window, /* cycleLog */ 0, maxDist, src);
&ldmState->window, /* cycleLog */ 0, maxDist, chunkStart);
ZSTD_ldm_reduceTable(ldmState->hashTable, ldmHSize, correction);
/* invalidate dictionaries on overflow correction */
ldmState->loadedDictEnd = 0;
}
/* 2. We enforce the maximum offset allowed.
*
@ -506,8 +474,14 @@ size_t ZSTD_ldm_generateSequences(
* TODO: * Test the chunk size.
* * Try invalidation after the sequence generation and test the
* the offset against maxDist directly.
*
* NOTE: Because of dictionaries + sequence splitting we MUST make sure
* that any offset used is valid at the END of the sequence, since it may
* be split into two sequences. This condition holds when using
* ZSTD_window_enforceMaxDist(), but if we move to checking offsets
* against maxDist directly, we'll have to carefully handle that case.
*/
ZSTD_window_enforceMaxDist(&ldmState->window, chunkEnd, maxDist, NULL, NULL);
ZSTD_window_enforceMaxDist(&ldmState->window, chunkEnd, maxDist, &ldmState->loadedDictEnd, NULL);
/* 3. Generate the sequences for the chunk, and get newLeftoverSize. */
newLeftoverSize = ZSTD_ldm_generateSequences_internal(
ldmState, sequences, params, chunkStart, chunkSize);
@ -593,7 +567,7 @@ size_t ZSTD_ldm_blockCompress(rawSeqStore_t* rawSeqStore,
void const* src, size_t srcSize)
{
const ZSTD_compressionParameters* const cParams = &ms->cParams;
unsigned const minMatch = cParams->searchLength;
unsigned const minMatch = cParams->minMatch;
ZSTD_blockCompressor const blockCompressor =
ZSTD_selectBlockCompressor(cParams->strategy, ZSTD_matchState_dictMode(ms));
/* Input bounds */
@ -615,14 +589,13 @@ size_t ZSTD_ldm_blockCompress(rawSeqStore_t* rawSeqStore,
if (sequence.offset == 0)
break;
assert(sequence.offset <= (1U << cParams->windowLog));
assert(ip + sequence.litLength + sequence.matchLength <= iend);
/* Fill tables for block compressor */
ZSTD_ldm_limitTableUpdate(ms, ip);
ZSTD_ldm_fillFastTables(ms, ip);
/* Run the block compressor */
DEBUGLOG(5, "calling block compressor on segment of size %u", sequence.litLength);
DEBUGLOG(5, "pos %u : calling block compressor on segment of size %u", (unsigned)(ip-istart), sequence.litLength);
{
size_t const newLitLength =
blockCompressor(ms, seqStore, rep, ip, sequence.litLength);
@ -632,7 +605,7 @@ size_t ZSTD_ldm_blockCompress(rawSeqStore_t* rawSeqStore,
rep[i] = rep[i-1];
rep[0] = sequence.offset;
/* Store the sequence */
ZSTD_storeSeq(seqStore, newLitLength, ip - newLitLength,
ZSTD_storeSeq(seqStore, newLitLength, ip - newLitLength, iend,
sequence.offset + ZSTD_REP_MOVE,
sequence.matchLength - MINMATCH);
ip += sequence.matchLength;

15
drivers/filesystems/btrfs/zstd/zstd_ldm.h
View file

@ -1,10 +1,11 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
* in the COPYING file in the root directory of this source tree).
* You may select, at your option, one of the above-listed licenses.
*/
#ifndef ZSTD_LDM_H
@ -21,7 +22,11 @@ extern "C" {
* Long distance matching
***************************************/
#define ZSTD_LDM_DEFAULT_WINDOW_LOG ZSTD_WINDOWLOG_DEFAULTMAX
#define ZSTD_LDM_DEFAULT_WINDOW_LOG ZSTD_WINDOWLOG_LIMIT_DEFAULT
void ZSTD_ldm_fillHashTable(
ldmState_t* state, const BYTE* ip,
const BYTE* iend, ldmParams_t const* params);
/**
* ZSTD_ldm_generateSequences():
@ -86,12 +91,8 @@ size_t ZSTD_ldm_getTableSize(ldmParams_t params);
*/
size_t ZSTD_ldm_getMaxNbSeq(ldmParams_t params, size_t maxChunkSize);
/** ZSTD_ldm_getTableSize() :
* Return prime8bytes^(minMatchLength-1) */
U64 ZSTD_ldm_getHashPower(U32 minMatchLength);
/** ZSTD_ldm_adjustParameters() :
* If the params->hashEveryLog is not set, set it to its default value based on
* If the params->hashRateLog is not set, set it to its default value based on
* windowLog and params->hashLog.
*
* Ensures that params->bucketSizeLog is <= params->hashLog (setting it to

382
drivers/filesystems/btrfs/zstd/zstd_opt.c
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Przemyslaw Skibinski, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Przemyslaw Skibinski, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -17,6 +17,8 @@
#define ZSTD_FREQ_DIV 4 /* log factor when using previous stats to init next stats */
#define ZSTD_MAX_PRICE (1<<30)
#define ZSTD_PREDEF_THRESHOLD 1024 /* if srcSize < ZSTD_PREDEF_THRESHOLD, symbols' cost is assumed static, directly determined by pre-defined distributions */
/*-*************************************
* Price functions for optimal parser
@ -52,14 +54,24 @@ MEM_STATIC U32 ZSTD_fracWeight(U32 rawStat)
return weight;
}
/* debugging function, @return price in bytes */
#if (DEBUGLEVEL>=2)
/* debugging function,
* @return price in bytes as fractional value
* for debug messages only */
MEM_STATIC double ZSTD_fCost(U32 price)
{
return (double)price / (BITCOST_MULTIPLIER*8);
}
#endif
static int ZSTD_compressedLiterals(optState_t const* const optPtr)
{
return optPtr->literalCompressionMode != ZSTD_lcm_uncompressed;
}
static void ZSTD_setBasePrices(optState_t* optPtr, int optLevel)
{
if (ZSTD_compressedLiterals(optPtr))
optPtr->litSumBasePrice = WEIGHT(optPtr->litSum, optLevel);
optPtr->litLengthSumBasePrice = WEIGHT(optPtr->litLengthSum, optLevel);
optPtr->matchLengthSumBasePrice = WEIGHT(optPtr->matchLengthSum, optLevel);
@ -67,34 +79,51 @@ static void ZSTD_setBasePrices(optState_t* optPtr, int optLevel)
}
static U32 ZSTD_downscaleStat(U32* table, U32 lastEltIndex, int malus)
/* ZSTD_downscaleStat() :
* reduce all elements in table by a factor 2^(ZSTD_FREQ_DIV+malus)
* return the resulting sum of elements */
static U32 ZSTD_downscaleStat(unsigned* table, U32 lastEltIndex, int malus)
{
U32 s, sum=0;
DEBUGLOG(5, "ZSTD_downscaleStat (nbElts=%u)", (unsigned)lastEltIndex+1);
assert(ZSTD_FREQ_DIV+malus > 0 && ZSTD_FREQ_DIV+malus < 31);
for (s=0; s<=lastEltIndex; s++) {
for (s=0; s<lastEltIndex+1; s++) {
table[s] = 1 + (table[s] >> (ZSTD_FREQ_DIV+malus));
sum += table[s];
}
return sum;
}
static void ZSTD_rescaleFreqs(optState_t* const optPtr,
/* ZSTD_rescaleFreqs() :
* if first block (detected by optPtr->litLengthSum == 0) : init statistics
* take hints from dictionary if there is one
* or init from zero, using src for literals stats, or flat 1 for match symbols
* otherwise downscale existing stats, to be used as seed for next block.
*/
static void
ZSTD_rescaleFreqs(optState_t* const optPtr,
const BYTE* const src, size_t const srcSize,
int optLevel)
int const optLevel)
{
int const compressedLiterals = ZSTD_compressedLiterals(optPtr);
DEBUGLOG(5, "ZSTD_rescaleFreqs (srcSize=%u)", (unsigned)srcSize);
optPtr->priceType = zop_dynamic;
if (optPtr->litLengthSum == 0) { /* first block : init */
if (srcSize <= 1024) /* heuristic */
if (srcSize <= ZSTD_PREDEF_THRESHOLD) { /* heuristic */
DEBUGLOG(5, "(srcSize <= ZSTD_PREDEF_THRESHOLD) => zop_predef");
optPtr->priceType = zop_predef;
}
assert(optPtr->symbolCosts != NULL);
if (optPtr->symbolCosts->huf.repeatMode == HUF_repeat_valid) { /* huffman table presumed generated by dictionary */
if (optPtr->symbolCosts->huf.repeatMode == HUF_repeat_valid) {
/* huffman table presumed generated by dictionary */
optPtr->priceType = zop_dynamic;
if (compressedLiterals) {
unsigned lit;
assert(optPtr->litFreq != NULL);
optPtr->litSum = 0;
{ unsigned lit;
for (lit=0; lit<=MaxLit; lit++) {
U32 const scaleLog = 11; /* scale to 2K */
U32 const bitCost = HUF_getNbBits(optPtr->symbolCosts->huf.CTable, lit);
@ -142,10 +171,11 @@ static void ZSTD_rescaleFreqs(optState_t* const optPtr,
} else { /* not a dictionary */
assert(optPtr->litFreq != NULL);
{ unsigned lit = MaxLit;
if (compressedLiterals) {
unsigned lit = MaxLit;
HIST_count_simple(optPtr->litFreq, &lit, src, srcSize); /* use raw first block to init statistics */
}
optPtr->litSum = ZSTD_downscaleStat(optPtr->litFreq, MaxLit, 1);
}
{ unsigned ll;
for (ll=0; ll<=MaxLL; ll++)
@ -169,6 +199,7 @@ static void ZSTD_rescaleFreqs(optState_t* const optPtr,
} else { /* new block : re-use previous statistics, scaled down */
if (compressedLiterals)
optPtr->litSum = ZSTD_downscaleStat(optPtr->litFreq, MaxLit, 1);
optPtr->litLengthSum = ZSTD_downscaleStat(optPtr->litLengthFreq, MaxLL, 0);
optPtr->matchLengthSum = ZSTD_downscaleStat(optPtr->matchLengthFreq, MaxML, 0);
@ -186,6 +217,10 @@ static U32 ZSTD_rawLiteralsCost(const BYTE* const literals, U32 const litLength,
int optLevel)
{
if (litLength == 0) return 0;
if (!ZSTD_compressedLiterals(optPtr))
return (litLength << 3) * BITCOST_MULTIPLIER; /* Uncompressed - 8 bytes per literal. */
if (optPtr->priceType == zop_predef)
return (litLength*6) * BITCOST_MULTIPLIER; /* 6 bit per literal - no statistic used */
@ -208,44 +243,12 @@ static U32 ZSTD_litLengthPrice(U32 const litLength, const optState_t* const optP
/* dynamic statistics */
{ U32 const llCode = ZSTD_LLcode(litLength);
return (LL_bits[llCode] * BITCOST_MULTIPLIER) + (optPtr->litLengthSumBasePrice - WEIGHT(optPtr->litLengthFreq[llCode], optLevel));
}
}
/* ZSTD_litLengthContribution() :
* @return ( cost(litlength) - cost(0) )
* this value can then be added to rawLiteralsCost()
* to provide a cost which is directly comparable to a match ending at same position */
static int ZSTD_litLengthContribution(U32 const litLength, const optState_t* const optPtr, int optLevel)
{
if (optPtr->priceType >= zop_predef) return WEIGHT(litLength, optLevel);
/* dynamic statistics */
{ U32 const llCode = ZSTD_LLcode(litLength);
int const contribution = (LL_bits[llCode] * BITCOST_MULTIPLIER)
+ WEIGHT(optPtr->litLengthFreq[0], optLevel) /* note: log2litLengthSum cancel out */
return (LL_bits[llCode] * BITCOST_MULTIPLIER)
+ optPtr->litLengthSumBasePrice
- WEIGHT(optPtr->litLengthFreq[llCode], optLevel);
#if 1
return contribution;
#else
return MAX(0, contribution); /* sometimes better, sometimes not ... */
#endif
}
}
/* ZSTD_literalsContribution() :
* creates a fake cost for the literals part of a sequence
* which can be compared to the ending cost of a match
* should a new match start at this position */
static int ZSTD_literalsContribution(const BYTE* const literals, U32 const litLength,
const optState_t* const optPtr,
int optLevel)
{
int const contribution = ZSTD_rawLiteralsCost(literals, litLength, optPtr, optLevel)
+ ZSTD_litLengthContribution(litLength, optPtr, optLevel);
return contribution;
}
/* ZSTD_getMatchPrice() :
* Provides the cost of the match part (offset + matchLength) of a sequence
* Must be combined with ZSTD_fullLiteralsCost() to get the full cost of a sequence.
@ -287,7 +290,8 @@ static void ZSTD_updateStats(optState_t* const optPtr,
U32 offsetCode, U32 matchLength)
{
/* literals */
{ U32 u;
if (ZSTD_compressedLiterals(optPtr)) {
U32 u;
for (u=0; u < litLength; u++)
optPtr->litFreq[literals[u]] += ZSTD_LITFREQ_ADD;
optPtr->litSum += litLength*ZSTD_LITFREQ_ADD;
@ -334,13 +338,15 @@ MEM_STATIC U32 ZSTD_readMINMATCH(const void* memPtr, U32 length)
/* Update hashTable3 up to ip (excluded)
Assumption : always within prefix (i.e. not within extDict) */
static U32 ZSTD_insertAndFindFirstIndexHash3 (ZSTD_matchState_t* ms, const BYTE* const ip)
static U32 ZSTD_insertAndFindFirstIndexHash3 (ZSTD_matchState_t* ms,
U32* nextToUpdate3,
const BYTE* const ip)
{
U32* const hashTable3 = ms->hashTable3;
U32 const hashLog3 = ms->hashLog3;
const BYTE* const base = ms->window.base;
U32 idx = ms->nextToUpdate3;
U32 const target = ms->nextToUpdate3 = (U32)(ip - base);
U32 idx = *nextToUpdate3;
U32 const target = (U32)(ip - base);
size_t const hash3 = ZSTD_hash3Ptr(ip, hashLog3);
assert(hashLog3 > 0);
@ -349,6 +355,7 @@ static U32 ZSTD_insertAndFindFirstIndexHash3 (ZSTD_matchState_t* ms, const BYTE*
idx++;
}
*nextToUpdate3 = target;
return hashTable3[hash3];
}
@ -385,7 +392,6 @@ static U32 ZSTD_insertBt1(
U32* largerPtr = smallerPtr + 1;
U32 dummy32; /* to be nullified at the end */
U32 const windowLow = ms->window.lowLimit;
U32 const matchLow = windowLow ? windowLow : 1;
U32 matchEndIdx = current+8+1;
size_t bestLength = 8;
U32 nbCompares = 1U << cParams->searchLog;
@ -401,7 +407,8 @@ static U32 ZSTD_insertBt1(
assert(ip <= iend-8); /* required for h calculation */
hashTable[h] = current; /* Update Hash Table */
while (nbCompares-- && (matchIndex >= matchLow)) {
assert(windowLow > 0);
while (nbCompares-- && (matchIndex >= windowLow)) {
U32* const nextPtr = bt + 2*(matchIndex & btMask);
size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */
assert(matchIndex < current);
@ -465,9 +472,11 @@ static U32 ZSTD_insertBt1(
} }
*smallerPtr = *largerPtr = 0;
if (bestLength > 384) return MIN(192, (U32)(bestLength - 384)); /* speed optimization */
{ U32 positions = 0;
if (bestLength > 384) positions = MIN(192, (U32)(bestLength - 384)); /* speed optimization */
assert(matchEndIdx > current + 8);
return matchEndIdx - (current + 8);
return MAX(positions, matchEndIdx - (current + 8));
}
}
FORCE_INLINE_TEMPLATE
@ -479,24 +488,33 @@ void ZSTD_updateTree_internal(
const BYTE* const base = ms->window.base;
U32 const target = (U32)(ip - base);
U32 idx = ms->nextToUpdate;
DEBUGLOG(5, "ZSTD_updateTree_internal, from %u to %u (dictMode:%u)",
DEBUGLOG(6, "ZSTD_updateTree_internal, from %u to %u (dictMode:%u)",
idx, target, dictMode);
while(idx < target)
idx += ZSTD_insertBt1(ms, base+idx, iend, mls, dictMode == ZSTD_extDict);
while(idx < target) {
U32 const forward = ZSTD_insertBt1(ms, base+idx, iend, mls, dictMode == ZSTD_extDict);
assert(idx < (U32)(idx + forward));
idx += forward;
}
assert((size_t)(ip - base) <= (size_t)(U32)(-1));
assert((size_t)(iend - base) <= (size_t)(U32)(-1));
ms->nextToUpdate = target;
}
void ZSTD_updateTree(ZSTD_matchState_t* ms, const BYTE* ip, const BYTE* iend) {
ZSTD_updateTree_internal(ms, ip, iend, ms->cParams.searchLength, ZSTD_noDict);
ZSTD_updateTree_internal(ms, ip, iend, ms->cParams.minMatch, ZSTD_noDict);
}
FORCE_INLINE_TEMPLATE
U32 ZSTD_insertBtAndGetAllMatches (
ZSTD_match_t* matches, /* store result (found matches) in this table (presumed large enough) */
ZSTD_matchState_t* ms,
U32* nextToUpdate3,
const BYTE* const ip, const BYTE* const iLimit, const ZSTD_dictMode_e dictMode,
U32 rep[ZSTD_REP_NUM], U32 const ll0,
ZSTD_match_t* matches, const U32 lengthToBeat, U32 const mls /* template */)
const U32 rep[ZSTD_REP_NUM],
U32 const ll0, /* tells if associated literal length is 0 or not. This value must be 0 or 1 */
const U32 lengthToBeat,
U32 const mls /* template */)
{
const ZSTD_compressionParameters* const cParams = &ms->cParams;
U32 const sufficient_len = MIN(cParams->targetLength, ZSTD_OPT_NUM -1);
@ -515,8 +533,8 @@ U32 ZSTD_insertBtAndGetAllMatches (
U32 const dictLimit = ms->window.dictLimit;
const BYTE* const dictEnd = dictBase + dictLimit;
const BYTE* const prefixStart = base + dictLimit;
U32 const btLow = btMask >= current ? 0 : current - btMask;
U32 const windowLow = ms->window.lowLimit;
U32 const btLow = (btMask >= current) ? 0 : current - btMask;
U32 const windowLow = ZSTD_getLowestMatchIndex(ms, current, cParams->windowLog);
U32 const matchLow = windowLow ? windowLow : 1;
U32* smallerPtr = bt + 2*(current&btMask);
U32* largerPtr = bt + 2*(current&btMask) + 1;
@ -542,6 +560,7 @@ U32 ZSTD_insertBtAndGetAllMatches (
DEBUGLOG(8, "ZSTD_insertBtAndGetAllMatches: current=%u", current);
/* check repCode */
assert(ll0 <= 1); /* necessarily 1 or 0 */
{ U32 const lastR = ZSTD_REP_NUM + ll0;
U32 repCode;
for (repCode = ll0; repCode < lastR; repCode++) {
@ -550,7 +569,10 @@ U32 ZSTD_insertBtAndGetAllMatches (
U32 repLen = 0;
assert(current >= dictLimit);
if (repOffset-1 /* intentional overflow, discards 0 and -1 */ < current-dictLimit) { /* equivalent to `current > repIndex >= dictLimit` */
if (ZSTD_readMINMATCH(ip, minMatch) == ZSTD_readMINMATCH(ip - repOffset, minMatch)) {
/* We must validate the repcode offset because when we're using a dictionary the
* valid offset range shrinks when the dictionary goes out of bounds.
*/
if ((repIndex >= windowLow) & (ZSTD_readMINMATCH(ip, minMatch) == ZSTD_readMINMATCH(ip - repOffset, minMatch))) {
repLen = (U32)ZSTD_count(ip+minMatch, ip+minMatch-repOffset, iLimit) + minMatch;
}
} else { /* repIndex < dictLimit || repIndex >= current */
@ -585,7 +607,7 @@ U32 ZSTD_insertBtAndGetAllMatches (
/* HC3 match finder */
if ((mls == 3) /*static*/ && (bestLength < mls)) {
U32 const matchIndex3 = ZSTD_insertAndFindFirstIndexHash3(ms, ip);
U32 const matchIndex3 = ZSTD_insertAndFindFirstIndexHash3(ms, nextToUpdate3, ip);
if ((matchIndex3 >= matchLow)
& (current - matchIndex3 < (1<<18)) /*heuristic : longer distance likely too expensive*/ ) {
size_t mlen;
@ -611,9 +633,7 @@ U32 ZSTD_insertBtAndGetAllMatches (
(ip+mlen == iLimit) ) { /* best possible length */
ms->nextToUpdate = current+1; /* skip insertion */
return 1;
}
}
}
} } }
/* no dictMatchState lookup: dicts don't have a populated HC3 table */
}
@ -621,19 +641,21 @@ U32 ZSTD_insertBtAndGetAllMatches (
while (nbCompares-- && (matchIndex >= matchLow)) {
U32* const nextPtr = bt + 2*(matchIndex & btMask);
size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */
const BYTE* match;
size_t matchLength = MIN(commonLengthSmaller, commonLengthLarger); /* guaranteed minimum nb of common bytes */
assert(current > matchIndex);
if ((dictMode == ZSTD_noDict) || (dictMode == ZSTD_dictMatchState) || (matchIndex+matchLength >= dictLimit)) {
assert(matchIndex+matchLength >= dictLimit); /* ensure the condition is correct when !extDict */
match = base + matchIndex;
if (matchIndex >= dictLimit) assert(memcmp(match, ip, matchLength) == 0); /* ensure early section of match is equal as expected */
matchLength += ZSTD_count(ip+matchLength, match+matchLength, iLimit);
} else {
match = dictBase + matchIndex;
assert(memcmp(match, ip, matchLength) == 0); /* ensure early section of match is equal as expected */
matchLength += ZSTD_count_2segments(ip+matchLength, match+matchLength, iLimit, dictEnd, prefixStart);
if (matchIndex+matchLength >= dictLimit)
match = base + matchIndex; /* prepare for match[matchLength] */
match = base + matchIndex; /* prepare for match[matchLength] read */
}
if (matchLength > bestLength) {
@ -718,24 +740,27 @@ U32 ZSTD_insertBtAndGetAllMatches (
FORCE_INLINE_TEMPLATE U32 ZSTD_BtGetAllMatches (
ZSTD_match_t* matches, /* store result (match found, increasing size) in this table */
ZSTD_matchState_t* ms,
U32* nextToUpdate3,
const BYTE* ip, const BYTE* const iHighLimit, const ZSTD_dictMode_e dictMode,
U32 rep[ZSTD_REP_NUM], U32 const ll0,
ZSTD_match_t* matches, U32 const lengthToBeat)
const U32 rep[ZSTD_REP_NUM],
U32 const ll0,
U32 const lengthToBeat)
{
const ZSTD_compressionParameters* const cParams = &ms->cParams;
U32 const matchLengthSearch = cParams->searchLength;
U32 const matchLengthSearch = cParams->minMatch;
DEBUGLOG(8, "ZSTD_BtGetAllMatches");
if (ip < ms->window.base + ms->nextToUpdate) return 0; /* skipped area */
ZSTD_updateTree_internal(ms, ip, iHighLimit, matchLengthSearch, dictMode);
switch(matchLengthSearch)
{
case 3 : return ZSTD_insertBtAndGetAllMatches(ms, ip, iHighLimit, dictMode, rep, ll0, matches, lengthToBeat, 3);
case 3 : return ZSTD_insertBtAndGetAllMatches(matches, ms, nextToUpdate3, ip, iHighLimit, dictMode, rep, ll0, lengthToBeat, 3);
default :
case 4 : return ZSTD_insertBtAndGetAllMatches(ms, ip, iHighLimit, dictMode, rep, ll0, matches, lengthToBeat, 4);
case 5 : return ZSTD_insertBtAndGetAllMatches(ms, ip, iHighLimit, dictMode, rep, ll0, matches, lengthToBeat, 5);
case 4 : return ZSTD_insertBtAndGetAllMatches(matches, ms, nextToUpdate3, ip, iHighLimit, dictMode, rep, ll0, lengthToBeat, 4);
case 5 : return ZSTD_insertBtAndGetAllMatches(matches, ms, nextToUpdate3, ip, iHighLimit, dictMode, rep, ll0, lengthToBeat, 5);
case 7 :
case 6 : return ZSTD_insertBtAndGetAllMatches(ms, ip, iHighLimit, dictMode, rep, ll0, matches, lengthToBeat, 6);
case 6 : return ZSTD_insertBtAndGetAllMatches(matches, ms, nextToUpdate3, ip, iHighLimit, dictMode, rep, ll0, lengthToBeat, 6);
}
}
@ -743,30 +768,6 @@ FORCE_INLINE_TEMPLATE U32 ZSTD_BtGetAllMatches (
/*-*******************************
* Optimal parser
*********************************/
typedef struct repcodes_s {
U32 rep[3];
} repcodes_t;
static repcodes_t ZSTD_updateRep(U32 const rep[3], U32 const offset, U32 const ll0)
{
repcodes_t newReps;
if (offset >= ZSTD_REP_NUM) { /* full offset */
newReps.rep[2] = rep[1];
newReps.rep[1] = rep[0];
newReps.rep[0] = offset - ZSTD_REP_MOVE;
} else { /* repcode */
U32 const repCode = offset + ll0;
if (repCode > 0) { /* note : if repCode==0, no change */
U32 const currentOffset = (repCode==ZSTD_REP_NUM) ? (rep[0] - 1) : rep[repCode];
newReps.rep[2] = (repCode >= 2) ? rep[1] : rep[2];
newReps.rep[1] = rep[0];
newReps.rep[0] = currentOffset;
} else { /* repCode == 0 */
memcpy(&newReps, rep, sizeof(newReps));
}
}
return newReps;
}
static U32 ZSTD_totalLen(ZSTD_optimal_t sol)
@ -774,12 +775,30 @@ static U32 ZSTD_totalLen(ZSTD_optimal_t sol)
return sol.litlen + sol.mlen;
}
#if 0 /* debug */
static void
listStats(const U32* table, int lastEltID)
{
int const nbElts = lastEltID + 1;
int enb;
for (enb=0; enb < nbElts; enb++) {
(void)table;
/* RAWLOG(2, "%3i:%3i, ", enb, table[enb]); */
RAWLOG(2, "%4i,", table[enb]);
}
RAWLOG(2, " \n");
}
#endif
FORCE_INLINE_TEMPLATE size_t
ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
seqStore_t* seqStore,
U32 rep[ZSTD_REP_NUM],
const void* src, size_t srcSize,
const int optLevel, const ZSTD_dictMode_e dictMode)
const int optLevel,
const ZSTD_dictMode_e dictMode)
{
optState_t* const optStatePtr = &ms->opt;
const BYTE* const istart = (const BYTE*)src;
@ -792,16 +811,17 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
const ZSTD_compressionParameters* const cParams = &ms->cParams;
U32 const sufficient_len = MIN(cParams->targetLength, ZSTD_OPT_NUM -1);
U32 const minMatch = (cParams->searchLength == 3) ? 3 : 4;
U32 const minMatch = (cParams->minMatch == 3) ? 3 : 4;
U32 nextToUpdate3 = ms->nextToUpdate;
ZSTD_optimal_t* const opt = optStatePtr->priceTable;
ZSTD_match_t* const matches = optStatePtr->matchTable;
ZSTD_optimal_t lastSequence;
/* init */
DEBUGLOG(5, "ZSTD_compressBlock_opt_generic");
DEBUGLOG(5, "ZSTD_compressBlock_opt_generic: current=%u, prefix=%u, nextToUpdate=%u",
(U32)(ip - base), ms->window.dictLimit, ms->nextToUpdate);
assert(optLevel <= 2);
ms->nextToUpdate3 = ms->nextToUpdate;
ZSTD_rescaleFreqs(optStatePtr, (const BYTE*)src, srcSize, optLevel);
ip += (ip==prefixStart);
@ -812,19 +832,24 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
/* find first match */
{ U32 const litlen = (U32)(ip - anchor);
U32 const ll0 = !litlen;
U32 const nbMatches = ZSTD_BtGetAllMatches(ms, ip, iend, dictMode, rep, ll0, matches, minMatch);
U32 const nbMatches = ZSTD_BtGetAllMatches(matches, ms, &nextToUpdate3, ip, iend, dictMode, rep, ll0, minMatch);
if (!nbMatches) { ip++; continue; }
/* initialize opt[0] */
{ U32 i ; for (i=0; i<ZSTD_REP_NUM; i++) opt[0].rep[i] = rep[i]; }
opt[0].mlen = 0; /* means is_a_literal */
opt[0].litlen = litlen;
opt[0].price = ZSTD_literalsContribution(anchor, litlen, optStatePtr, optLevel);
/* We don't need to include the actual price of the literals because
* it is static for the duration of the forward pass, and is included
* in every price. We include the literal length to avoid negative
* prices when we subtract the previous literal length.
*/
opt[0].price = ZSTD_litLengthPrice(litlen, optStatePtr, optLevel);
/* large match -> immediate encoding */
{ U32 const maxML = matches[nbMatches-1].len;
U32 const maxOffset = matches[nbMatches-1].off;
DEBUGLOG(6, "found %u matches of maxLength=%u and maxOffCode=%u at cPos=%u => start new serie",
DEBUGLOG(6, "found %u matches of maxLength=%u and maxOffCode=%u at cPos=%u => start new series",
nbMatches, maxML, maxOffset, (U32)(ip-prefixStart));
if (maxML > sufficient_len) {
@ -848,7 +873,6 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
for (matchNb = 0; matchNb < nbMatches; matchNb++) {
U32 const offset = matches[matchNb].off;
U32 const end = matches[matchNb].len;
repcodes_t const repHistory = ZSTD_updateRep(rep, offset, ll0);
for ( ; pos <= end ; pos++ ) {
U32 const matchPrice = ZSTD_getMatchPrice(offset, pos, optStatePtr, optLevel);
U32 const sequencePrice = literalsPrice + matchPrice;
@ -858,8 +882,6 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
opt[pos].off = offset;
opt[pos].litlen = litlen;
opt[pos].price = sequencePrice;
ZSTD_STATIC_ASSERT(sizeof(opt[pos].rep) == sizeof(repHistory));
memcpy(opt[pos].rep, &repHistory, sizeof(repHistory));
} }
last_pos = pos-1;
}
@ -886,7 +908,6 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
opt[cur].off = 0;
opt[cur].litlen = litlen;
opt[cur].price = price;
memcpy(opt[cur].rep, opt[cur-1].rep, sizeof(opt[cur].rep));
} else {
DEBUGLOG(7, "cPos:%zi==rPos:%u : literal would cost more (%.2f>%.2f) (hist:%u,%u,%u)",
inr-istart, cur, ZSTD_fCost(price), ZSTD_fCost(opt[cur].price),
@ -894,6 +915,21 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
}
}
/* Set the repcodes of the current position. We must do it here
* because we rely on the repcodes of the 2nd to last sequence being
* correct to set the next chunks repcodes during the backward
* traversal.
*/
ZSTD_STATIC_ASSERT(sizeof(opt[cur].rep) == sizeof(repcodes_t));
assert(cur >= opt[cur].mlen);
if (opt[cur].mlen != 0) {
U32 const prev = cur - opt[cur].mlen;
repcodes_t newReps = ZSTD_updateRep(opt[prev].rep, opt[cur].off, opt[cur].litlen==0);
memcpy(opt[cur].rep, &newReps, sizeof(repcodes_t));
} else {
memcpy(opt[cur].rep, opt[cur - 1].rep, sizeof(repcodes_t));
}
/* last match must start at a minimum distance of 8 from oend */
if (inr > ilimit) continue;
@ -909,7 +945,7 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
U32 const litlen = (opt[cur].mlen == 0) ? opt[cur].litlen : 0;
U32 const previousPrice = opt[cur].price;
U32 const basePrice = previousPrice + ZSTD_litLengthPrice(0, optStatePtr, optLevel);
U32 const nbMatches = ZSTD_BtGetAllMatches(ms, inr, iend, dictMode, opt[cur].rep, ll0, matches, minMatch);
U32 const nbMatches = ZSTD_BtGetAllMatches(matches, ms, &nextToUpdate3, inr, iend, dictMode, opt[cur].rep, ll0, minMatch);
U32 matchNb;
if (!nbMatches) {
DEBUGLOG(7, "rPos:%u : no match found", cur);
@ -934,7 +970,6 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
/* set prices using matches found at position == cur */
for (matchNb = 0; matchNb < nbMatches; matchNb++) {
U32 const offset = matches[matchNb].off;
repcodes_t const repHistory = ZSTD_updateRep(opt[cur].rep, offset, ll0);
U32 const lastML = matches[matchNb].len;
U32 const startML = (matchNb>0) ? matches[matchNb-1].len+1 : minMatch;
U32 mlen;
@ -954,8 +989,6 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
opt[pos].off = offset;
opt[pos].litlen = litlen;
opt[pos].price = price;
ZSTD_STATIC_ASSERT(sizeof(opt[pos].rep) == sizeof(repHistory));
memcpy(opt[pos].rep, &repHistory, sizeof(repHistory));
} else {
DEBUGLOG(7, "rPos:%u (ml=%2u) => new price is worse (%.2f>=%.2f)",
pos, mlen, ZSTD_fCost(price), ZSTD_fCost(opt[pos].price));
@ -971,6 +1004,17 @@ ZSTD_compressBlock_opt_generic(ZSTD_matchState_t* ms,
_shortestPath: /* cur, last_pos, best_mlen, best_off have to be set */
assert(opt[0].mlen == 0);
/* Set the next chunk's repcodes based on the repcodes of the beginning
* of the last match, and the last sequence. This avoids us having to
* update them while traversing the sequences.
*/
if (lastSequence.mlen != 0) {
repcodes_t reps = ZSTD_updateRep(opt[cur].rep, lastSequence.off, lastSequence.litlen==0);
memcpy(rep, &reps, sizeof(reps));
} else {
memcpy(rep, opt[cur].rep, sizeof(repcodes_t));
}
{ U32 const storeEnd = cur + 1;
U32 storeStart = storeEnd;
U32 seqPos = cur;
@ -999,7 +1043,7 @@ _shortestPath: /* cur, last_pos, best_mlen, best_off have to be set */
U32 const offCode = opt[storePos].off;
U32 const advance = llen + mlen;
DEBUGLOG(6, "considering seq starting at %zi, llen=%u, mlen=%u",
anchor - istart, llen, mlen);
anchor - istart, (unsigned)llen, (unsigned)mlen);
if (mlen==0) { /* only literals => must be last "sequence", actually starting a new stream of sequences */
assert(storePos == storeEnd); /* must be last sequence */
@ -1007,33 +1051,18 @@ _shortestPath: /* cur, last_pos, best_mlen, best_off have to be set */
continue; /* will finish */
}
/* repcodes update : like ZSTD_updateRep(), but update in place */
if (offCode >= ZSTD_REP_NUM) { /* full offset */
rep[2] = rep[1];
rep[1] = rep[0];
rep[0] = offCode - ZSTD_REP_MOVE;
} else { /* repcode */
U32 const repCode = offCode + (llen==0);
if (repCode) { /* note : if repCode==0, no change */
U32 const currentOffset = (repCode==ZSTD_REP_NUM) ? (rep[0] - 1) : rep[repCode];
if (repCode >= 2) rep[2] = rep[1];
rep[1] = rep[0];
rep[0] = currentOffset;
} }
assert(anchor + llen <= iend);
ZSTD_updateStats(optStatePtr, llen, anchor, offCode, mlen);
ZSTD_storeSeq(seqStore, llen, anchor, offCode, mlen-MINMATCH);
ZSTD_storeSeq(seqStore, llen, anchor, iend, offCode, mlen-MINMATCH);
anchor += advance;
ip = anchor;
} }
ZSTD_setBasePrices(optStatePtr, optLevel);
}
} /* while (ip < ilimit) */
/* Return the last literals size */
return iend - anchor;
return (size_t)(iend - anchor);
}
@ -1047,11 +1076,11 @@ size_t ZSTD_compressBlock_btopt(
/* used in 2-pass strategy */
static U32 ZSTD_upscaleStat(U32* table, U32 lastEltIndex, int bonus)
static U32 ZSTD_upscaleStat(unsigned* table, U32 lastEltIndex, int bonus)
{
U32 s, sum=0;
assert(ZSTD_FREQ_DIV+bonus > 0);
for (s=0; s<=lastEltIndex; s++) {
assert(ZSTD_FREQ_DIV+bonus >= 0);
for (s=0; s<lastEltIndex+1; s++) {
table[s] <<= ZSTD_FREQ_DIV+bonus;
table[s]--;
sum += table[s];
@ -1062,10 +1091,44 @@ static U32 ZSTD_upscaleStat(U32* table, U32 lastEltIndex, int bonus)
/* used in 2-pass strategy */
MEM_STATIC void ZSTD_upscaleStats(optState_t* optPtr)
{
if (ZSTD_compressedLiterals(optPtr))
optPtr->litSum = ZSTD_upscaleStat(optPtr->litFreq, MaxLit, 0);
optPtr->litLengthSum = ZSTD_upscaleStat(optPtr->litLengthFreq, MaxLL, 1);
optPtr->matchLengthSum = ZSTD_upscaleStat(optPtr->matchLengthFreq, MaxML, 1);
optPtr->offCodeSum = ZSTD_upscaleStat(optPtr->offCodeFreq, MaxOff, 1);
optPtr->litLengthSum = ZSTD_upscaleStat(optPtr->litLengthFreq, MaxLL, 0);
optPtr->matchLengthSum = ZSTD_upscaleStat(optPtr->matchLengthFreq, MaxML, 0);
optPtr->offCodeSum = ZSTD_upscaleStat(optPtr->offCodeFreq, MaxOff, 0);
}
/* ZSTD_initStats_ultra():
* make a first compression pass, just to seed stats with more accurate starting values.
* only works on first block, with no dictionary and no ldm.
* this function cannot error, hence its contract must be respected.
*/
static void
ZSTD_initStats_ultra(ZSTD_matchState_t* ms,
seqStore_t* seqStore,
U32 rep[ZSTD_REP_NUM],
const void* src, size_t srcSize)
{
U32 tmpRep[ZSTD_REP_NUM]; /* updated rep codes will sink here */
memcpy(tmpRep, rep, sizeof(tmpRep));
DEBUGLOG(4, "ZSTD_initStats_ultra (srcSize=%zu)", srcSize);
assert(ms->opt.litLengthSum == 0); /* first block */
assert(seqStore->sequences == seqStore->sequencesStart); /* no ldm */
assert(ms->window.dictLimit == ms->window.lowLimit); /* no dictionary */
assert(ms->window.dictLimit - ms->nextToUpdate <= 1); /* no prefix (note: intentional overflow, defined as 2-complement) */
ZSTD_compressBlock_opt_generic(ms, seqStore, tmpRep, src, srcSize, 2 /*optLevel*/, ZSTD_noDict); /* generate stats into ms->opt*/
/* invalidate first scan from history */
ZSTD_resetSeqStore(seqStore);
ms->window.base -= srcSize;
ms->window.dictLimit += (U32)srcSize;
ms->window.lowLimit = ms->window.dictLimit;
ms->nextToUpdate = ms->window.dictLimit;
/* re-inforce weight of collected statistics */
ZSTD_upscaleStats(&ms->opt);
}
size_t ZSTD_compressBlock_btultra(
@ -1073,33 +1136,34 @@ size_t ZSTD_compressBlock_btultra(
const void* src, size_t srcSize)
{
DEBUGLOG(5, "ZSTD_compressBlock_btultra (srcSize=%zu)", srcSize);
#if 0
/* 2-pass strategy (disabled)
return ZSTD_compressBlock_opt_generic(ms, seqStore, rep, src, srcSize, 2 /*optLevel*/, ZSTD_noDict);
}
size_t ZSTD_compressBlock_btultra2(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
const void* src, size_t srcSize)
{
U32 const current = (U32)((const BYTE*)src - ms->window.base);
DEBUGLOG(5, "ZSTD_compressBlock_btultra2 (srcSize=%zu)", srcSize);
/* 2-pass strategy:
* this strategy makes a first pass over first block to collect statistics
* and seed next round's statistics with it.
* After 1st pass, function forgets everything, and starts a new block.
* Consequently, this can only work if no data has been previously loaded in tables,
* aka, no dictionary, no prefix, no ldm preprocessing.
* The compression ratio gain is generally small (~0.5% on first block),
* the cost is 2x cpu time on first block. */
assert(srcSize <= ZSTD_BLOCKSIZE_MAX);
if ( (ms->opt.litLengthSum==0) /* first block */
&& (seqStore->sequences == seqStore->sequencesStart) /* no ldm */
&& (ms->window.dictLimit == ms->window.lowLimit) ) { /* no dictionary */
U32 tmpRep[ZSTD_REP_NUM];
DEBUGLOG(5, "ZSTD_compressBlock_btultra: first block: collecting statistics");
assert(ms->nextToUpdate >= ms->window.dictLimit
&& ms->nextToUpdate <= ms->window.dictLimit + 1);
memcpy(tmpRep, rep, sizeof(tmpRep));
ZSTD_compressBlock_opt_generic(ms, seqStore, tmpRep, src, srcSize, 2 /*optLevel*/, ZSTD_noDict); /* generate stats into ms->opt*/
ZSTD_resetSeqStore(seqStore);
/* invalidate first scan from history */
ms->window.base -= srcSize;
ms->window.dictLimit += (U32)srcSize;
ms->window.lowLimit = ms->window.dictLimit;
ms->nextToUpdate = ms->window.dictLimit;
ms->nextToUpdate3 = ms->window.dictLimit;
/* re-inforce weight of collected statistics */
ZSTD_upscaleStats(&ms->opt);
&& (ms->window.dictLimit == ms->window.lowLimit) /* no dictionary */
&& (current == ms->window.dictLimit) /* start of frame, nothing already loaded nor skipped */
&& (srcSize > ZSTD_PREDEF_THRESHOLD)
) {
ZSTD_initStats_ultra(ms, seqStore, rep, src, srcSize);
}
#endif
return ZSTD_compressBlock_opt_generic(ms, seqStore, rep, src, srcSize, 2 /*optLevel*/, ZSTD_noDict);
}
@ -1130,3 +1194,7 @@ size_t ZSTD_compressBlock_btultra_extDict(
{
return ZSTD_compressBlock_opt_generic(ms, seqStore, rep, src, srcSize, 2 /*optLevel*/, ZSTD_extDict);
}
/* note : no btultra2 variant for extDict nor dictMatchState,
* because btultra2 is not meant to work with dictionaries
* and is only specific for the first block (no prefix) */

10
drivers/filesystems/btrfs/zstd/zstd_opt.h
View file

@ -1,5 +1,5 @@
/*
* Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
* Copyright (c) 2016-2020, Yann Collet, Facebook, Inc.
* All rights reserved.
*
* This source code is licensed under both the BSD-style license (found in the
@ -26,6 +26,10 @@ size_t ZSTD_compressBlock_btopt(
size_t ZSTD_compressBlock_btultra(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize);
size_t ZSTD_compressBlock_btultra2(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize);
size_t ZSTD_compressBlock_btopt_dictMatchState(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
@ -41,6 +45,10 @@ size_t ZSTD_compressBlock_btultra_extDict(
ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM],
void const* src, size_t srcSize);
/* note : no btultra2 variant for extDict nor dictMatchState,
* because btultra2 is not meant to work with dictionaries
* and is only specific for the first block (no prefix) */
#if defined (__cplusplus)
}
#endif

9
sdk/lib/fslib/btrfslib/CMakeLists.txt
View file

@ -9,14 +9,11 @@ list(APPEND SOURCE
${REACTOS_SOURCE_DIR}/drivers/filesystems/btrfs/xxhash.c
btrfslib.c)
if(ARCH STREQUAL "i386")
list(APPEND ASM_SOURCE ${REACTOS_SOURCE_DIR}/drivers/filesystems/btrfs/crc32c-x86.S)
elseif(ARCH STREQUAL "amd64")
list(APPEND ASM_SOURCE ${REACTOS_SOURCE_DIR}/drivers/filesystems/btrfs/crc32c-amd64.S)
if((ARCH STREQUAL "i386") OR (ARCH STREQUAL "amd64"))
list(APPEND ASM_SOURCE ${REACTOS_SOURCE_DIR}/drivers/filesystems/btrfs/crc32c.S)
add_asm_files(btrfs_asm ${ASM_SOURCE})
endif()
add_asm_files(btrfs_asm ${ASM_SOURCE})
add_library(btrfslib ${SOURCE} ${btrfs_asm})
target_compile_definitions(btrfslib PRIVATE _USRDLL)
add_dependencies(btrfslib psdk)