Missing factor of \pi in the argument of the Green function for the top-level mesh potential calculation.

Improvements to GIZMO implementation

See merge request !524

Make the brute-force tests use the correct cosmological parameters and also call the chemistry functions.

Bert Vandenbroucke authored Apr 10, 2018 and Matthieu Schaller committed Apr 10, 2018

Removed unused isothermal Riemann solver. Added generic input and output debug checks for all remaining Riemann solvers. Made sure HLLC Riemann solver works for 0 input pressure. Made sure all quantities are set to vacuum if density/mass or pressure/energy is zero.

# Conflicts:
#   src/Makefile.am

Ignore differences in dh/dt for the active pair test if the norm of the value is less than 3e-5. This is close enough to zero.

Split the EoS definition into multiple files for clarity

See merge request !523

Re-ordered some operations in the scalar versions of SPH schemes to take advantage of the faster rsqrt calls.

Compute the correct particle ID in the gravity force checks also for the SPH particles and star particles.

Various issues, now reproduces the plots we produced before, also have a new mpiactive column