Sink drift

In this merge request, I am implementing the drift/kick/timestep tasks for the sink particles.

I have done 2 tests. The first is the example EAGLE_low_z/EAGLE_6 where I turned on all the EAGLE's physics (swift --eagle) and was able to run until the end (17 steps). The second one is IsolatedGalaxy/IsolatedGalaxy_sink. I have changed it in order to have some gas particles (otherwise SWIFT is not able to run with sink particles due to the way we construct the tasks). All the sink particles are falling into the galaxy without any strange behavior (e.g. visible cell not updated). This is only a visual test, I did not check against any analytical solution but it should be fine as they evolve according to the gpart.

The merge request might seem a bit large, but in most of the files, there is only 1-5 lines that are modified. If it is too large, I can try to reduce it a bit.