optimize more integer cases in floatscan; comment the whole procedure

[musl] / src / stdlib / abs.c

int abs(int a)
{
        return a>0 ? a : -a;
}