ext-cryptopp/misc.h

811 lines
19 KiB
C++

#ifndef CRYPTOPP_MISC_H
#define CRYPTOPP_MISC_H
#include "cryptlib.h"
#include "smartptr.h"
#ifdef INTEL_INTRINSICS
#include <stdlib.h>
#endif
NAMESPACE_BEGIN(CryptoPP)
// ************** compile-time assertion ***************
template <bool b>
struct CompileAssert
{
static char dummy[2*b-1];
};
#define CRYPTOPP_COMPILE_ASSERT(assertion) CRYPTOPP_COMPILE_ASSERT_INSTANCE(assertion, __LINE__)
#if defined(CRYPTOPP_EXPORTS) || defined(CRYPTOPP_IMPORTS)
#define CRYPTOPP_COMPILE_ASSERT_INSTANCE(assertion, instance)
#else
#define CRYPTOPP_COMPILE_ASSERT_INSTANCE(assertion, instance) static CompileAssert<(assertion)> CRYPTOPP_ASSERT_JOIN(cryptopp_assert_, instance)
#endif
#define CRYPTOPP_ASSERT_JOIN(X, Y) CRYPTOPP_DO_ASSERT_JOIN(X, Y)
#define CRYPTOPP_DO_ASSERT_JOIN(X, Y) X##Y
// ************** misc classes ***************
class CRYPTOPP_DLL Empty
{
};
template <class BASE1, class BASE2>
class CRYPTOPP_NO_VTABLE TwoBases : public BASE1, public BASE2
{
};
template <class BASE1, class BASE2, class BASE3>
class CRYPTOPP_NO_VTABLE ThreeBases : public BASE1, public BASE2, public BASE3
{
};
template <class T>
class ObjectHolder
{
protected:
T m_object;
};
class NotCopyable
{
public:
NotCopyable() {}
private:
NotCopyable(const NotCopyable &);
void operator=(const NotCopyable &);
};
// ************** misc functions ***************
// can't use std::min or std::max in MSVC60 or Cygwin 1.1.0
template <class _Tp> inline const _Tp& STDMIN(const _Tp& __a, const _Tp& __b)
{
return __b < __a ? __b : __a;
}
template <class _Tp> inline const _Tp& STDMAX(const _Tp& __a, const _Tp& __b)
{
return __a < __b ? __b : __a;
}
#define RETURN_IF_NONZERO(x) unsigned int returnedValue = x; if (returnedValue) return returnedValue
// this version of the macro is fastest on Pentium 3 and Pentium 4 with MSVC 6 SP5 w/ Processor Pack
#define GETBYTE(x, y) (unsigned int)byte((x)>>(8*(y)))
// these may be faster on other CPUs/compilers
// #define GETBYTE(x, y) (unsigned int)(((x)>>(8*(y)))&255)
// #define GETBYTE(x, y) (((byte *)&(x))[y])
CRYPTOPP_DLL unsigned int Parity(unsigned long);
CRYPTOPP_DLL unsigned int BytePrecision(unsigned long);
CRYPTOPP_DLL unsigned int BitPrecision(unsigned long);
CRYPTOPP_DLL unsigned long Crop(unsigned long, unsigned int size);
inline unsigned int BitsToBytes(unsigned int bitCount)
{
return ((bitCount+7)/(8));
}
inline unsigned int BytesToWords(unsigned int byteCount)
{
return ((byteCount+WORD_SIZE-1)/WORD_SIZE);
}
inline unsigned int BitsToWords(unsigned int bitCount)
{
return ((bitCount+WORD_BITS-1)/(WORD_BITS));
}
inline unsigned int BitsToDwords(unsigned int bitCount)
{
return ((bitCount+2*WORD_BITS-1)/(2*WORD_BITS));
}
CRYPTOPP_DLL void xorbuf(byte *buf, const byte *mask, unsigned int count);
CRYPTOPP_DLL void xorbuf(byte *output, const byte *input, const byte *mask, unsigned int count);
template <class T>
inline bool IsPowerOf2(T n)
{
return n > 0 && (n & (n-1)) == 0;
}
template <class T1, class T2>
inline T2 ModPowerOf2(T1 a, T2 b)
{
assert(IsPowerOf2(b));
return T2(a) & (b-1);
}
template <class T>
inline T RoundDownToMultipleOf(T n, T m)
{
return n - (IsPowerOf2(m) ? ModPowerOf2(n, m) : (n%m));
}
template <class T>
inline T RoundUpToMultipleOf(T n, T m)
{
return RoundDownToMultipleOf(n+m-1, m);
}
template <class T>
inline unsigned int GetAlignment(T *dummy=NULL) // VC60 workaround
{
#if (_MSC_VER >= 1300)
return __alignof(T);
#elif defined(__GNUC__)
return __alignof__(T);
#else
return sizeof(T);
#endif
}
inline bool IsAlignedOn(const void *p, unsigned int alignment)
{
return IsPowerOf2(alignment) ? ModPowerOf2((size_t)p, alignment) == 0 : (size_t)p % alignment == 0;
}
template <class T>
inline bool IsAligned(const void *p, T *dummy=NULL) // VC60 workaround
{
return IsAlignedOn(p, GetAlignment<T>());
}
#ifdef IS_LITTLE_ENDIAN
typedef LittleEndian NativeByteOrder;
#else
typedef BigEndian NativeByteOrder;
#endif
inline ByteOrder GetNativeByteOrder()
{
return NativeByteOrder::ToEnum();
}
inline bool NativeByteOrderIs(ByteOrder order)
{
return order == GetNativeByteOrder();
}
template <class T> // can't use <sstream> because GCC 2.95.2 doesn't have it
std::string IntToString(T a, unsigned int base = 10)
{
if (a == 0)
return "0";
bool negate = false;
if (a < 0)
{
negate = true;
a = 0-a; // VC .NET does not like -a
}
std::string result;
while (a > 0)
{
T digit = a % base;
result = char((digit < 10 ? '0' : ('a' - 10)) + digit) + result;
a /= base;
}
if (negate)
result = "-" + result;
return result;
}
template <class T1, class T2>
inline T1 SaturatingSubtract(T1 a, T2 b)
{
CRYPTOPP_COMPILE_ASSERT_INSTANCE(T1(-1)>0, 0); // T1 is unsigned type
CRYPTOPP_COMPILE_ASSERT_INSTANCE(T2(-1)>0, 1); // T2 is unsigned type
return T1((a > b) ? (a - b) : 0);
}
template <class T>
inline CipherDir GetCipherDir(const T &obj)
{
return obj.IsForwardTransformation() ? ENCRYPTION : DECRYPTION;
}
template <class T>
struct NewObject
{
T* operator()() const {return new T;}
};
// This function safely initializes a static object in a multithreaded environment without using locks.
// It may leak memory when two threads try to initialize the static object at the same time
// but this should be acceptable since each static object is only initialized once per session.
template <class T, class F = NewObject<T>, int instance=0>
class Singleton
{
public:
Singleton(F objectFactory = F()) : m_objectFactory(objectFactory) {}
// VC60 workaround: use "..." to prevent this function from being inlined
const T & Ref(...) const;
private:
F m_objectFactory;
};
template <class T, class F, int instance>
const T & Singleton<T, F, instance>::Ref(...) const
{
static simple_ptr<T> s_pObject;
static char s_objectState = 0;
retry:
switch (s_objectState)
{
case 0:
s_objectState = 1;
try
{
s_pObject.m_p = m_objectFactory();
}
catch(...)
{
s_objectState = 0;
throw;
}
s_objectState = 2;
break;
case 1:
goto retry;
default:
break;
}
return *s_pObject.m_p;
}
// ************** rotate functions ***************
template <class T> inline T rotlFixed(T x, unsigned int y)
{
assert(y < sizeof(T)*8);
return (x<<y) | (x>>(sizeof(T)*8-y));
}
template <class T> inline T rotrFixed(T x, unsigned int y)
{
assert(y < sizeof(T)*8);
return (x>>y) | (x<<(sizeof(T)*8-y));
}
template <class T> inline T rotlVariable(T x, unsigned int y)
{
assert(y < sizeof(T)*8);
return (x<<y) | (x>>(sizeof(T)*8-y));
}
template <class T> inline T rotrVariable(T x, unsigned int y)
{
assert(y < sizeof(T)*8);
return (x>>y) | (x<<(sizeof(T)*8-y));
}
template <class T> inline T rotlMod(T x, unsigned int y)
{
y %= sizeof(T)*8;
return (x<<y) | (x>>(sizeof(T)*8-y));
}
template <class T> inline T rotrMod(T x, unsigned int y)
{
y %= sizeof(T)*8;
return (x>>y) | (x<<(sizeof(T)*8-y));
}
#ifdef INTEL_INTRINSICS
#pragma intrinsic(_lrotl, _lrotr)
template<> inline word32 rotlFixed<word32>(word32 x, unsigned int y)
{
assert(y < 32);
return y ? _lrotl(x, y) : x;
}
template<> inline word32 rotrFixed<word32>(word32 x, unsigned int y)
{
assert(y < 32);
return y ? _lrotr(x, y) : x;
}
template<> inline word32 rotlVariable<word32>(word32 x, unsigned int y)
{
assert(y < 32);
return _lrotl(x, y);
}
template<> inline word32 rotrVariable<word32>(word32 x, unsigned int y)
{
assert(y < 32);
return _lrotr(x, y);
}
template<> inline word32 rotlMod<word32>(word32 x, unsigned int y)
{
return _lrotl(x, y);
}
template<> inline word32 rotrMod<word32>(word32 x, unsigned int y)
{
return _lrotr(x, y);
}
#endif // #ifdef INTEL_INTRINSICS
#ifdef PPC_INTRINSICS
template<> inline word32 rotlFixed<word32>(word32 x, unsigned int y)
{
assert(y < 32);
return y ? __rlwinm(x,y,0,31) : x;
}
template<> inline word32 rotrFixed<word32>(word32 x, unsigned int y)
{
assert(y < 32);
return y ? __rlwinm(x,32-y,0,31) : x;
}
template<> inline word32 rotlVariable<word32>(word32 x, unsigned int y)
{
assert(y < 32);
return (__rlwnm(x,y,0,31));
}
template<> inline word32 rotrVariable<word32>(word32 x, unsigned int y)
{
assert(y < 32);
return (__rlwnm(x,32-y,0,31));
}
template<> inline word32 rotlMod<word32>(word32 x, unsigned int y)
{
return (__rlwnm(x,y,0,31));
}
template<> inline word32 rotrMod<word32>(word32 x, unsigned int y)
{
return (__rlwnm(x,32-y,0,31));
}
#endif // #ifdef PPC_INTRINSICS
// ************** endian reversal ***************
template <class T>
inline unsigned int GetByte(ByteOrder order, T value, unsigned int index)
{
if (order == LITTLE_ENDIAN_ORDER)
return GETBYTE(value, index);
else
return GETBYTE(value, sizeof(T)-index-1);
}
inline byte ByteReverse(byte value)
{
return value;
}
inline word16 ByteReverse(word16 value)
{
return rotlFixed(value, 8U);
}
inline word32 ByteReverse(word32 value)
{
#ifdef PPC_INTRINSICS
// PPC: load reverse indexed instruction
return (word32)__lwbrx(&value,0);
#elif defined(FAST_ROTATE)
// 5 instructions with rotate instruction, 9 without
return (rotrFixed(value, 8U) & 0xff00ff00) | (rotlFixed(value, 8U) & 0x00ff00ff);
#else
// 6 instructions with rotate instruction, 8 without
value = ((value & 0xFF00FF00) >> 8) | ((value & 0x00FF00FF) << 8);
return rotlFixed(value, 16U);
#endif
}
#ifdef WORD64_AVAILABLE
inline word64 ByteReverse(word64 value)
{
#ifdef CRYPTOPP_SLOW_WORD64
return (word64(ByteReverse(word32(value))) << 32) | ByteReverse(word32(value>>32));
#else
value = ((value & W64LIT(0xFF00FF00FF00FF00)) >> 8) | ((value & W64LIT(0x00FF00FF00FF00FF)) << 8);
value = ((value & W64LIT(0xFFFF0000FFFF0000)) >> 16) | ((value & W64LIT(0x0000FFFF0000FFFF)) << 16);
return rotlFixed(value, 32U);
#endif
}
#endif
inline byte BitReverse(byte value)
{
value = ((value & 0xAA) >> 1) | ((value & 0x55) << 1);
value = ((value & 0xCC) >> 2) | ((value & 0x33) << 2);
return rotlFixed(value, 4);
}
inline word16 BitReverse(word16 value)
{
value = ((value & 0xAAAA) >> 1) | ((value & 0x5555) << 1);
value = ((value & 0xCCCC) >> 2) | ((value & 0x3333) << 2);
value = ((value & 0xF0F0) >> 4) | ((value & 0x0F0F) << 4);
return ByteReverse(value);
}
inline word32 BitReverse(word32 value)
{
value = ((value & 0xAAAAAAAA) >> 1) | ((value & 0x55555555) << 1);
value = ((value & 0xCCCCCCCC) >> 2) | ((value & 0x33333333) << 2);
value = ((value & 0xF0F0F0F0) >> 4) | ((value & 0x0F0F0F0F) << 4);
return ByteReverse(value);
}
#ifdef WORD64_AVAILABLE
inline word64 BitReverse(word64 value)
{
#ifdef CRYPTOPP_SLOW_WORD64
return (word64(BitReverse(word32(value))) << 32) | BitReverse(word32(value>>32));
#else
value = ((value & W64LIT(0xAAAAAAAAAAAAAAAA)) >> 1) | ((value & W64LIT(0x5555555555555555)) << 1);
value = ((value & W64LIT(0xCCCCCCCCCCCCCCCC)) >> 2) | ((value & W64LIT(0x3333333333333333)) << 2);
value = ((value & W64LIT(0xF0F0F0F0F0F0F0F0)) >> 4) | ((value & W64LIT(0x0F0F0F0F0F0F0F0F)) << 4);
return ByteReverse(value);
#endif
}
#endif
template <class T>
inline T BitReverse(T value)
{
if (sizeof(T) == 1)
return (T)BitReverse((byte)value);
else if (sizeof(T) == 2)
return (T)BitReverse((word16)value);
else if (sizeof(T) == 4)
return (T)BitReverse((word32)value);
else
{
#ifdef WORD64_AVAILABLE
assert(sizeof(T) == 8);
return (T)BitReverse((word64)value);
#else
assert(false);
return 0;
#endif
}
}
template <class T>
inline T ConditionalByteReverse(ByteOrder order, T value)
{
return NativeByteOrderIs(order) ? value : ByteReverse(value);
}
template <class T>
void ByteReverse(T *out, const T *in, unsigned int byteCount)
{
assert(byteCount % sizeof(T) == 0);
unsigned int count = byteCount/sizeof(T);
for (unsigned int i=0; i<count; i++)
out[i] = ByteReverse(in[i]);
}
template <class T>
inline void ConditionalByteReverse(ByteOrder order, T *out, const T *in, unsigned int byteCount)
{
if (!NativeByteOrderIs(order))
ByteReverse(out, in, byteCount);
else if (in != out)
memcpy(out, in, byteCount);
}
template <class T>
inline void GetUserKey(ByteOrder order, T *out, unsigned int outlen, const byte *in, unsigned int inlen)
{
const unsigned int U = sizeof(T);
assert(inlen <= outlen*U);
memcpy(out, in, inlen);
memset((byte *)out+inlen, 0, outlen*U-inlen);
ConditionalByteReverse(order, out, out, RoundUpToMultipleOf(inlen, U));
}
inline byte UnalignedGetWordNonTemplate(ByteOrder order, const byte *block, byte*)
{
return block[0];
}
inline word16 UnalignedGetWordNonTemplate(ByteOrder order, const byte *block, word16*)
{
return (order == BIG_ENDIAN_ORDER)
? block[1] | (block[0] << 8)
: block[0] | (block[1] << 8);
}
inline word32 UnalignedGetWordNonTemplate(ByteOrder order, const byte *block, word32*)
{
return (order == BIG_ENDIAN_ORDER)
? word32(block[3]) | (word32(block[2]) << 8) | (word32(block[1]) << 16) | (word32(block[0]) << 24)
: word32(block[0]) | (word32(block[1]) << 8) | (word32(block[2]) << 16) | (word32(block[3]) << 24);
}
#ifdef WORD64_AVAILABLE
inline word64 UnalignedGetWordNonTemplate(ByteOrder order, const byte *block, word64*)
{
return (order == BIG_ENDIAN_ORDER)
?
(word64(block[7]) |
(word64(block[6]) << 8) |
(word64(block[5]) << 16) |
(word64(block[4]) << 24) |
(word64(block[3]) << 32) |
(word64(block[2]) << 40) |
(word64(block[1]) << 48) |
(word64(block[0]) << 56))
:
(word64(block[0]) |
(word64(block[1]) << 8) |
(word64(block[2]) << 16) |
(word64(block[3]) << 24) |
(word64(block[4]) << 32) |
(word64(block[5]) << 40) |
(word64(block[6]) << 48) |
(word64(block[7]) << 56));
}
#endif
template <class T>
inline T UnalignedGetWord(ByteOrder order, const byte *block, T*dummy=NULL)
{
return UnalignedGetWordNonTemplate(order, block, dummy);
}
inline void UnalignedPutWord(ByteOrder order, byte *block, byte value, const byte *xorBlock = NULL)
{
block[0] = xorBlock ? (value ^ xorBlock[0]) : value;
}
inline void UnalignedPutWord(ByteOrder order, byte *block, word16 value, const byte *xorBlock = NULL)
{
if (order == BIG_ENDIAN_ORDER)
{
block[0] = GETBYTE(value, 1);
block[1] = GETBYTE(value, 0);
}
else
{
block[0] = GETBYTE(value, 0);
block[1] = GETBYTE(value, 1);
}
if (xorBlock)
{
block[0] ^= xorBlock[0];
block[1] ^= xorBlock[1];
}
}
inline void UnalignedPutWord(ByteOrder order, byte *block, word32 value, const byte *xorBlock = NULL)
{
if (order == BIG_ENDIAN_ORDER)
{
block[0] = GETBYTE(value, 3);
block[1] = GETBYTE(value, 2);
block[2] = GETBYTE(value, 1);
block[3] = GETBYTE(value, 0);
}
else
{
block[0] = GETBYTE(value, 0);
block[1] = GETBYTE(value, 1);
block[2] = GETBYTE(value, 2);
block[3] = GETBYTE(value, 3);
}
if (xorBlock)
{
block[0] ^= xorBlock[0];
block[1] ^= xorBlock[1];
block[2] ^= xorBlock[2];
block[3] ^= xorBlock[3];
}
}
#ifdef WORD64_AVAILABLE
inline void UnalignedPutWord(ByteOrder order, byte *block, word64 value, const byte *xorBlock = NULL)
{
if (order == BIG_ENDIAN_ORDER)
{
block[0] = GETBYTE(value, 7);
block[1] = GETBYTE(value, 6);
block[2] = GETBYTE(value, 5);
block[3] = GETBYTE(value, 4);
block[4] = GETBYTE(value, 3);
block[5] = GETBYTE(value, 2);
block[6] = GETBYTE(value, 1);
block[7] = GETBYTE(value, 0);
}
else
{
block[0] = GETBYTE(value, 0);
block[1] = GETBYTE(value, 1);
block[2] = GETBYTE(value, 2);
block[3] = GETBYTE(value, 3);
block[4] = GETBYTE(value, 4);
block[5] = GETBYTE(value, 5);
block[6] = GETBYTE(value, 6);
block[7] = GETBYTE(value, 7);
}
if (xorBlock)
{
block[0] ^= xorBlock[0];
block[1] ^= xorBlock[1];
block[2] ^= xorBlock[2];
block[3] ^= xorBlock[3];
block[4] ^= xorBlock[4];
block[5] ^= xorBlock[5];
block[6] ^= xorBlock[6];
block[7] ^= xorBlock[7];
}
}
#endif
template <class T>
inline T GetWord(bool assumeAligned, ByteOrder order, const byte *block)
{
if (assumeAligned)
{
assert(IsAligned<T>(block));
return ConditionalByteReverse(order, *reinterpret_cast<const T *>(block));
}
else
return UnalignedGetWord<T>(order, block);
}
template <class T>
inline void GetWord(bool assumeAligned, ByteOrder order, T &result, const byte *block)
{
result = GetWord<T>(assumeAligned, order, block);
}
template <class T>
inline void PutWord(bool assumeAligned, ByteOrder order, byte *block, T value, const byte *xorBlock = NULL)
{
if (assumeAligned)
{
assert(IsAligned<T>(block));
if (xorBlock)
*reinterpret_cast<T *>(block) = ConditionalByteReverse(order, value) ^ *reinterpret_cast<const T *>(xorBlock);
else
*reinterpret_cast<T *>(block) = ConditionalByteReverse(order, value);
}
else
UnalignedPutWord(order, block, value, xorBlock);
}
template <class T, class B, bool A=true>
class GetBlock
{
public:
GetBlock(const void *block)
: m_block((const byte *)block) {}
template <class U>
inline GetBlock<T, B, A> & operator()(U &x)
{
CRYPTOPP_COMPILE_ASSERT(sizeof(U) >= sizeof(T));
x = GetWord<T>(A, B::ToEnum(), m_block);
m_block += sizeof(T);
return *this;
}
private:
const byte *m_block;
};
template <class T, class B, bool A=true>
class PutBlock
{
public:
PutBlock(const void *xorBlock, void *block)
: m_xorBlock((const byte *)xorBlock), m_block((byte *)block) {}
template <class U>
inline PutBlock<T, B, A> & operator()(U x)
{
PutWord(A, B::ToEnum(), m_block, (T)x, m_xorBlock);
m_block += sizeof(T);
if (m_xorBlock)
m_xorBlock += sizeof(T);
return *this;
}
private:
const byte *m_xorBlock;
byte *m_block;
};
template <class T, class B, bool A=true>
struct BlockGetAndPut
{
// function needed because of C++ grammatical ambiguity between expression-statements and declarations
static inline GetBlock<T, B, A> Get(const void *block) {return GetBlock<T, B, A>(block);}
typedef PutBlock<T, B, A> Put;
};
template <class T>
std::string WordToString(T value, ByteOrder order = BIG_ENDIAN_ORDER)
{
if (!NativeByteOrderIs(order))
value = ByteReverse(value);
return std::string((char *)&value, sizeof(value));
}
template <class T>
T StringToWord(const std::string &str, ByteOrder order = BIG_ENDIAN_ORDER)
{
T value = 0;
memcpy(&value, str.data(), STDMIN(sizeof(value), str.size()));
return NativeByteOrderIs(order) ? value : ByteReverse(value);
}
// ************** help remove warning on g++ ***************
template <bool overflow> struct SafeShifter;
template<> struct SafeShifter<true>
{
template <class T>
static inline T RightShift(T value, unsigned int bits)
{
return 0;
}
template <class T>
static inline T LeftShift(T value, unsigned int bits)
{
return 0;
}
};
template<> struct SafeShifter<false>
{
template <class T>
static inline T RightShift(T value, unsigned int bits)
{
return value >> bits;
}
template <class T>
static inline T LeftShift(T value, unsigned int bits)
{
return value << bits;
}
};
template <unsigned int bits, class T>
inline T SafeRightShift(T value)
{
return SafeShifter<(bits>=(8*sizeof(T)))>::RightShift(value, bits);
}
template <unsigned int bits, class T>
inline T SafeLeftShift(T value)
{
return SafeShifter<(bits>=(8*sizeof(T)))>::LeftShift(value, bits);
}
NAMESPACE_END
#endif // MISC_H