Make more pow functions accept fmpz

[wbhart]

For example n_powmod2 could have a version that accepts an fmpz. (There's an adhoc version in one of the nmod_powmod_blah functions.)

In general, any function whose output would be small for very large exponent should have an fmpz version.

This has already been done for the nmod_poly_powmod_blah and fmpz_mod_poly_powmod_blah functions.

[tthsqe12]

I think I also added {fmpz|fmpq}_pow_fmpz, which return a bool, so functions that could fail are also useful.

[albinahlback]

I think this only applies for finite rings or finite precision rings? Although we already have fmpz_pow_fmpz, I don't think it makes a lot of sense for such modules where it is only a very finite set of elements (only units, right?) that allows for exponents larger than $2^{\mathtt{FLINT\_BITS}}$.