Pseudorandom seeding across OSs

Torbjorn Granlund tg at
Sat Aug 27 12:08:06 CEST 2011

James Wanless <james at> writes:

  > The most important part is to include a self-contained test case that
  > illustrates the problem.
  Yes, I was afraid you might say that ( :) ). If I get inspired I'll
  look into it (I guess it's something I'll need to investigate for GMP- 
  ECPP now anyway) - but I make no promises!
It is possible to write GMP application code that results in all sorts
of machine dependent behaviour.  Some GMP interfaces are inherently type
size dependent.  The _ui functions when passed a 'long', might gove
different results due to truncation in the caller.

The fact that you have written some code that exhibits apparent type
size dependent behaviour does not in any way indicate a bug in GMP.

My suggestion is that you isolate the problem, and then, if the problem
is to some extent related to GMP and you need to discuss the problem,
you make a post here.


More information about the gmp-discuss mailing list