segfaults when using init_iters > 1

[zingale]

If I run xrb_mixed in 2-d with inputs_2d.rprox with maestro.init_iter=3 it crashes. Changing this to 1 makes it work.

[zingale]

it also works with init_iter=2 curiously

With init_iter = 3, we get:

Timestep 1 starts with TIME = 0 DT = 8.675232218e-08

Cell Count:
Level 0, 36864 cells
inner sponge: r_sp      , r_tp      : 3, 111
<<< STEP 1 : react state >>>
<<< STEP 2 : make w0 >>>
<<< STEP 3 : create MAC velocities >>>
MLMG: Initial rhs               = 5.455142391e+21
MLMG: Initial residual (resid0) = 5.455142391e+21
MLMG: Final Iter. 7 resid, resid/bnorm = 3.658298163e+11, 6.706146056e-11
MLMG: Timers: Solve = 0.010994001 Iter = 0.010767289 Bottom = 0.006200959
<<< STEP 4 : advect base >>>
            :  density_advance >>>
            :   tracer_advance >>>
            : enthalpy_advance >>>
<<< STEP 4a: thermal conduct >>>
<<< STEP 5 : react state >>>
Erroneous arithmetic operation
See Backtrace.1 file for details

[zingale]

I don't see this behavior in reacting_bubble

[zingale]

curiously, it still crashes at the same place if I run with maestro.do_burning=0

[zingale]

okay, curiously when I run it under valgrind, it runs fine. No crash :(
(it's out to timestep 9)

[zingale]

when running with

./Maestro2d.gnu.MPI.ex inputs_2d.rprox maestro.do_burning=0 maestro.evolve_base_state=0

we have a divide by zero in MaestroMakew0 because p0_new_arr is zero, and it is in the denominator for delta_chi_w0

[zingale]

somehow the new pressure is really getting messed up, in MaestroAdvance.cpp, but I am not sure why