[FLASH-BUGS] running flash with runtime error checking

[Peter Woitke]

Dear developers,

searching for my own errors, I switched on runtime error checking
(using ifort -g -CB -CU -CV -traceback). I got various error
messages. You may want to have a look at them:

1) source/mesh/amr/paramesh2.0/quadratic_spherical/amr_prolong_gen_unk_fun.F90,
   line 334: idz is not assigned any value (zero by default), so the
   dBaseGetCoords-call makes no sense and dy() gets arbitrary data.
   I propose the following solution:

   !call dBaseGetCoords(idz,  iYcoord, isg, dy)        ! changed PW
   #if N_DIM >= 2
     dyc = y(3) - y(1)
   #else
     dyc = 0.0
   #endif
   dy(:) = y(2)-y(1)                                   ! changed PW,

   is that ok?

2) forrtl: severe (408): fort: (5): Variable OUTPUT_DIRECTORY has
   substring starting point 0 which is less than one
   Image       PC        Routine           Line    Source
   flash2   081AEEE6  get_output_filena      56    get_output_filename.F90
   flash2   0818566D  checkpoint_wr_        273    checkpoint.F90
   flash2   08182E63  output_initial_       124    output_initial.F90
   flash2   08094E73  init_flash_           301    init_flash.F90
   flash2   08093A9D  MAIN__                 59    flash.F90
   (location source/io/get_output_filename.F90)
   (no solution proposed, ignored)

3) forrtl: severe (408): fort: (2): Subscript #2 of the array UNK1BLK has
   value 477 which is greater than the upper bound of 16
   Image       PC        Routine            Line   Source
   flash2   0811F841  modulehydrosweep_      384   hydro_sweep.F90
   flash2   080BA1A9  hydro_mp_hydro_3d      109   hydro.F90
   flash2   080ADE0B  evolve_                 71   evolve.F90
   flash2   08093F23  MAIN__                  96   flash.F90
   (location source/hydro/explicit/split/hydro_sweep.F90)
   (no solution proposed, ignored)

I would appreciate any comment

Peter Woitke