Better support for loading restart files

[aprilnovak]

Describe the bug I have a case with a restart file created for N = 7. If I load that restart file for a case with either N = 5 or N = 7, things run okay. But if I instead load it into a new case with N = 3, I get:

 Reading checkpoint data
       FILE:/home/anovak/ldrd/cheng2009/multiphysics/coupled/from_bebop/restart.fld
      15 ERROR: :    24219648    14130816             mfi_getv a
      35 ERROR: :    24219648    14130816             mfi_getv a
      37 ERROR: :    24219648    14130816             mfi_getv a
      39 ERROR: :    24219648    14130816             mfi_getv a
      10 ERROR: :    24219648    14130816             mfi_getv a
      16 ERROR: :    24219648    14130816             mfi_getv a
      19 ERROR: :    24219648    14130816             mfi_getv a
      28 ERROR: :    24219648    14130816             mfi_getv a
      29 ERROR: :    24219648    14130816             mfi_getv a
      13 ERROR: :    24219648    14130816             mfi_getv a
      18 ERROR: :    24219648    14130816             mfi_getv a
      20 ERROR: :    24219648    14130816             mfi_getv a
      21 ERROR: :    24219648    14130816             mfi_getv a
      38 ERROR: :    24219648    14130816             mfi_getv a
      14 ERROR: :    24219648    14130816             mfi_getv a
      24 ERROR: :    24219648    14130816             mfi_getv a
      27 ERROR: :    24219648    14130816             mfi_getv a
      30 ERROR: :    24219648    14130816             mfi_getv a
       2 ERROR: :    24219648    14130816             mfi_getv a
       8 ERROR: :    24219648    14130816             mfi_getv a
       9 ERROR: :    24219648    14130816             mfi_getv a
       0 ERROR: :    24219648    14130816             mfi_getv a
EXIT: lim_chk problem.    24219648

Expected behavior NekRS needs better support for mixing polynomial orders between restart files and current cases. At the very least, please throw an error to the user because the error messages above are not useful in identify the cause.

[stgeke]

This is a nek5000 thing. Can you please report it here: https://github.com/Nek5000/Nek5000/issues

[aprilnovak]

of course! Apologies for the misunderstanding