自己实现几个基本函数

void _Assert(char* strFile, unsigned line)
{
    fflush(stdout);
    fprintf(stderr, "\nAssert Failed %s, line:%u\n", strFile, line);
    fflush(stderr);
    abort();
}

#ifdef _DEBUG
#define MY_ASSERT(f) \
    if (f)          \
    NULL;           \
    else            \
    _Assert(__FILE__, __LINE__);
#else
#define MY_ASSERT(f)   NULL
#endif

void* my_memcpy(void* pvTo, void* pvFrom, size_t size)
{
    byte* pbTo = (byte*)pvTo;
    byte* pbFrom = (byte*)pvFrom;
    MY_ASSERT(pbTo != NULL && pbFrom != NULL);
    MY_ASSERT(pbTo >= pbFrom + size || pbFrom >= pbTo + size);

    while(size-- > 0)
        *pbTo++ = *pbFrom++;
    return (void*)pbTo;
}

char* my_strdup(char* str)
{
    MY_ASSERT(str != NULL);
    char* strNew = (char*)malloc(strlen(str) + 1);
    if (strNew == NULL)
        return NULL;
    //strcpy（str1，str2）时，str2会把末尾的'\0’也给复制到str1中且覆盖一个str1字符;
    strcpy(strNew, str);
    return strNew;
}

void* my_memset1(void* pv, byte b, size_t size)
{
    MY_ASSERT(pv != NULL);
    byte* pb = (byte*)pv;
    while (size-- > 0)
        *pb++ = b;
    return pv;
}

// 假定long占4字节，一字节为8位
long* long_fill(void* pv, long l, size_t size)
{
    MY_ASSERT(pv != NULL);
    long* pl = (long*)pv;
    while (size > 0)
    {
        *pl = l;
        size -= 4;
        pl += 4;
    }
    return (long*)pv;
}

size_t threshold = 4;
void* my_memset2(void* pv, byte b, size_t size)
{
    byte* pb = (byte*)pv;
    if (size > threshold)
    {
        unsigned long l;
        l = (b << 8) | b;
        l = (l << 16) | l;
        pb = (byte*)long_fill((long*)pb, l, size / 4);
        size = size % 4;
    }

    while (size -- >0)
        *(pb++) = b;
    return pv;
}

struct Prim
{
    int pos;
    int index;
};

Prim vecPrim[N];

bool binarySearch(int pos, int& index)
{
    int first = 0;
    int last = N - 1;
    int mid = 0;
    while (first <= last)
    {
        mid = (first + last) / 2;
        if (vecPrim[mid].pos < pos)
            first = mid + 1;
        else if (vecPrim[mid].pos > pos)
            last = mid - 1;
        else
        {
            index = vecPrim[mid].pos;
            return true;
        }
    }
    return false;
}

template<class T>
int highestBitSet(T input)
{
    register int result;
    assert(input);        // zero is invalid input!
    assert(sizeof(T)==4); // 32bit data only!
    _asm bsr eax, input
    _asm mov result, eax
    return result;
}


template<class T>
T nearestPowerOfTwo(T input)
{  
    // the least possible power-of-two value is 1
    if (input <= 1) return 1;

    int highestBit = highestBitSet(input);
    int roundingTest = input & (1<< (highestBit-1)); 
    if (roundingTest) 
        ++highestBit;
    return static_cast<T>(1<<highestBit);
}

inline int lowestBitSet (unsigned short input)
{
    register int result;
    assert(input);       // zero is invalid input!
    _asm mov dx, input   // copy into a 32bit reg
    _asm and edx, 0xffff // keep only the bits we want
    _asm bsf eax, edx    // perform the scan
    _asm mov result, eax
    return result;
}