The int `nviter` is not increased in the OUT.DAT

[ajpearcey]

The number of iterations of the vmcon optimisation is given by the int nviter. It gives useful info in diagnosing how the solver is failing to find a solution. I do not think it is updating correctly currently. For instance the large-tokamak scenario produces the OUT.DAT:

 VMCON error flag                                                         (ifail)                           1     
 Number of iteration variables                                            (nvar)                           45     
 Number of constraints (total)                                            (neqns+nineqns)                  26     
 Optimisation switch                                                      (ioptimz)                         1     
 Figure of merit switch                                                   (minmax)                          1     
 Square root of the sum of squares of the constraint residuals            (sqsumsq)                 5.381E-05  OP 
 VMCON convergence parameter                                              (convergence_parameter)   0.000E+00  OP 
 Normalised objective function                                            (norm_objf)               1.600E+00  OP 
 Number of VMCON iterations                                               (nviter)                          0  OP 

ifail = 1 and nviter = 0 should be inconsistent. This is making it hard to work out whether the solver really does give up on the first QSP or not.

ref: @timothy-nunn @jonmaddock

[je-cook]

same for the MFILE see #2937

[timothy-nunn]

The convergence parameter is also 0, for the same reason as the nviter: new VMCON does not feedback like old VMCON did. This should be easy to fix using the callback in new VMCON