Cleaned up some equations in Gizmo's hydro_iact and hydro_gradients. Moved the gradient prediction to before the velocity boost.

Moved Gizmo to GizmoMFV, and changed gizmo configuration option to gizmo-mfv. Made sure all unit tests compile again with gizmo-mfv.

Removed time step from flux computation. Removed (untested) flux limiter and unused variables in hydro_part. Discovered and fixed a possible bug in Gizmo particle velocities treatment in flux computation.

Partially restored old time step criterion. Isolated flux limiter into a separate file and made it configurable. Removed some superfluous gravity variables.

Moved around the energy gravity update so that it uses the old momentum. Removed a superfluous time step in mflux.

Fixed bug in signal velocity computation. Added relative velocity of fluid w.r.t. particle velocity to timestep criterion.

Implemented the changes in ls that prevent gizmo crashes in cases where the mesh is seriously distorted

Added some extra isothermal eos flags in Gizmo/hydro.h. Made sure gpart mass is updated when conserved mass changes. Fixed a number of possible issues with the isothermal Riemann solver.

Added some extra checks to correctly handle vacuum + some debug checks. 2D Noh test now seems to work with moving particles.

Streamlined GIZMO_FIX_PARTICLES usage (tried to add parameter, but this does not work, as space_init has no access to the hydro_properties). Moved conserved variable initialization to hydro_first_init_part, and removed it from hydro_convert_quantities. Added some sanity checks to maintain positive density and pressure.

Cleaned up Gizmo fixes. Added a hydro_timestep_extra method and added an empty version for all hydro schemes.

Replaced all calls to fminf and fmaxf by the macros that was added for the integer min/max functions.

Broke one of my own rules by doing too much in a single commit: implemented runner_extra_ghost, moved gradient calculations for both gradient types to new gradient loop, fixed bug in engine.c involving the number of links (which is larger with an extra loop), streamlined gradient methods, documented most GIZMO specific files. Second loop works.