fix float_t and double_t defs on x86 when -mfpmath=sse -msse2 is used

[musl] / src / math / ldexpl.c

#include <math.h>

long double ldexpl(long double x, int n)
{
        return scalbnl(x, n);
}