Hi, is there any reason why there is no mpz_[cft]div_2exp function in GMP? Could it be more efficient than the following? mpz_tdiv_q_2exp (q, n, b); mpz_tdiv_r_2exp (r, n, b); Paul Zimmermann