d3q Code

Code fixed for new XC interface, although:
1. I'm not sure it works with libxc
2. I've given up for the moment it being spin-polarized ready

Improved restart of d3_tk.x by setting a checkpoint after the SMA calculation and another one after the initial Af

Small changes

QHA example

doc

Fixes equired for new functional modules, the fixing method is brutal and stupid performance-wise in a couple of places, but it really doesn ot matter because it is a minimal part of this code

More cleanup

Little clean ups

Fix for code not working in parallel any more

Code taken up to level with libxc vectorization, this part is a bit tricky I hope everything works ok (the examples do)