Missing check on DmRngeXm values for MPA calculation in metals

[matteodalessio98]

In MPA calculations for a metallic system, when the first value of DmRngeXm is zero, for instance

% DmRngeXm
 0.000000 | 2.721138 |         eV    # [Xm] Damping range
% 

the calculation may fail without errors in the logs. In my case I saw these lines in the slurm output:
** On entry to ZGEBAL parameter number 3 had an illegal value

The reason is the zero imaginary part in the first pole, which likely causes NaN in the MPA solver routine.

A check on this value could be introduced to cleanly stop the execution with a clear error message.

[giacomosesti]

Also for metallic system, now referring to PPA calculations, it could be useful to allow to control Im(ω) for the calculations of Xp at Re(ω)=0. Currently, Im(ω) is set to 0 and in the calculations of Xp it is kept a negligible Re(ω), but for metallic systems one may prefer a finite damping through Im(ω) to avoid potential numerical problems.

[attacc]

OK I added a check for DmRngeXm in a branch, I will send it for pull request soon