`--cosmo infinity` calculation fails for HS-

[TyBalduf]

Describe the bug Running HS- with --cosmo infinity (or --cpcmx <any_solvent> since this will have do a cosmo infinity solvent run first), I get this output in the SCC.

 iter      E             dE          RMSdq      gap      omega  full diag
   1     -4.0645745 -0.406457E+01  0.682E+00    5.96       0.0  T
   2     -4.0657628 -0.118828E-02  0.374E+00    5.60       1.0  T
Upps!

I found the error message here in the code and looking at the subroutine gjinv, it seems like can only fail if the matrix its working with is singular. This is based on the (commented out) warning messages in the subroutine.

There isn't likely much that can be done about a singular matrix (maybe some kind of pseudo-inverse?), but it would be useful to enable these warning messages to clarify what has gone wrong with the calculation.

Additional context Input: hs_debug_sdf.txt Output: new_hs.txt