Torbjörn Granlund tg at
Fri Nov 13 13:46:14 UTC 2015

I haven't followed this discussion in detail, but perhaps
gmp/nextprime.c (i.e., nextprime.c at the top-level) is a good start?

This sieve uses a tiny sieving table (512 bytes) and thus will behave
unusually poorly when sieving for primes p where sqrt(p) >> 512.  That
is fixable, e.g., by keeping offsets in a binary minheap.

Please encrypt, key id 0xC8601622

More information about the gmp-discuss mailing list