TODO


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28

GMP has mpz_divexact(q,n,d), we should have zdiv_exact(q,n,d).
It uses optimised division algorithm that requires that d|n.

Add zsets_radix
Add zstr_radix

Can zmodpowu and zmodpow be improved using some other algorithm?
Is it worth implementing precomputed optimal
  addition-chain exponentiation in zpowu?

Test big endian
Test always having .used > 0 for zero
  Test negative/non-negative instead of sign
Test long .sign
Test always having .chars % 4 == 0

Test optimisation of zmul:
  bc = [(Hb * Hc) << (m2 << 1)]
     + [(Hb * Hc) << m2]
     - [(Hb - Lb)(Hc - Lc) << m2]
     + [(Lb * Lc) << m2]
     + (Lb * Lc)

Would zmul be faster if we split only one of the
factors until they are both approximately the same
size?

Add entropy test for zrand.