Re: divappr interface

2018-05-09 Thread paul zimmermann
Niels, > From: ni...@lysator.liu.se (Niels Möller) > Date: Fri, 27 Apr 2018 22:28:39 +0200 > > ni...@lysator.liu.se (Niels Möller) writes: > > > Once we reach qn = dn - 1, keep looping to produce quotient limbs, but > > also discard one limb of dp in each interation, until we in the final

Re: divappr interface

2018-04-27 Thread Niels Möller
ni...@lysator.liu.se (Niels Möller) writes: > Once we reach qn = dn - 1, keep looping to produce quotient limbs, but > also discard one limb of dp in each interation, until we in the final > iteration have qn = 1, qn+2 = 3 numerator limbs, and qn+1 = 2 divisor > limbs, i.e., a single udiv_qr_3by2

divappr interface

2018-04-27 Thread Niels Möller
Hi, after the discussion on div_q, I had another look at divappr, or more precisely, the mpn_sbpi1_divappr_q function. It's not entirely obvious from the implementation which limbs of the {np, nn} argument are actually used. But I think it should access only the qn+2 upper limbs. For accuracy, i