Using library functions or bitwise operations? Which method is more efficient?

See more:

C++

Method 1

if ((size & (!(size & (size - 1)))) == size)//if size is exact power of 2
{
    alignedSegmentSize = size;
}
else
{
    while( size != 0)
    {
        size = size << 1;
        alignedSegmentSize = alignedSegmentSize + 1;
    }
    alignedSegmentSize = alignedSegmentSize << 1;
}

Method 2

double alignedSegment = log(static_cast<double>(size))/(log(TWO));
alignedSegment = (ceil((alignedSegment)));
int nextBufferSize = pow(2,alignedSegment);

I wanted to know which method will be more efficient to calculate the nearest power of two for the given value.


Thanks in advance