mod to symmetric range

Torbjorn Granlund tg at
Sun Sep 5 20:52:14 CEST 2010

Roman Pearce <rpearcea at> writes:

  I'm trying to efficiently compute r = a mod b with
  -(b-1)/2 <= r <= b/2.  Any suggestions?  All of my
  solutions move too much data around.
Is each b invvarint over several reductions?
Then pre-computing b/2 (using a 2exp function) will help.


More information about the gmp-discuss mailing list