Problem with gmp_randinit_set

[Pedro Gimeno]

Torbjörn Granlund wrote, On 2017-02-16 23:30:
> Perhaps we should expand testing wrt this?
> 
> We should check that seeding is effective, that gmp_randinit_set(b, a)
> make a and b behave the same including that the seed follows, and that
> further seeding affects both identically.  More?

I'd add a check for a few known values after some ~700 extractions with known seeds, to ensure cross-platform compatibility. Also, to try reseeding after taking every code path where the function pointers get set currently to the _NoSeed version, not just gmp_randinit_set.

I'd appreciate if someone takes care of this. I'm quite busy now.