求问关于光速乘的正确性

long long mul(long long a, long long b) {
	long long rt = a *b - ((long long)((long double)a / mod * b + 0.5)) * mod;
    return rt < 0 ? rt + mod : rt;
}