Linear response issue solved?

[Luis Alvarez-Thon]

Dear Dirac developers,

Thank you for the new version of DIRAC!

Regarding my previous thread about the linear response performance:

https://groups.google.com/u/1/g/dirac-users/c/3HNMgoDb-qs/m/PrR3goJ_CwAJ

Starting from DIRAC17 to DIRAC22 the calculations of current densities and related properties are taking too much time (three times slower). This is most notorious for production calculations when I must use larger basis sets. I THINK I have detected where the problem arises from. Look in the file:

src/london_direct.F90

in   subroutine xc_contrib_2el_lao, after the comment line

!   spin density contributions (xc kernel)

There is a new code:

    if (.not. dft_cfg_no_sdft) then

#ifdef VAR_MPI
      if (parcal) call dirac_parctl(XCINT_PAR)
#endif
      if (expped) then
!        s_b_lao and gs_b_lao export (used in fde)
         call integrate_xc(xc_mat_dim              = ntbas(0),        &
                           xc_nz                   = nz,              &
                           xc_dmat_0               = dmat,            &
                           xc_nr_dmat              = 0,               &
                           xc_nr_fmat              = max_nr_xyz_comp, &
                           xc_fmat                 = fmat,            &
                           xc_fmat_pg_sym          = isym_2el_lao,    &
                           xc_do_london_rhs_direct_der2 = .true.,     &
                           xc_london_direct_export = .true.)
      else
         call integrate_xc(xc_mat_dim              = ntbas(0),        &
                           xc_nz                   = nz,              &
                           xc_dmat_0               = dmat,            &
                           xc_nr_dmat              = 0,               &
                           xc_nr_fmat              = max_nr_xyz_comp, &
                           xc_fmat                 = fmat,            &
                           xc_fmat_pg_sym          = isym_2el_lao,    &
                           xc_do_london_rhs_direct_der2 = .true.)
      end if
    end if

Upon deactivation of  this part, everything works fine and I can reproduce my DIRAC2016 calculations, even a little faster that before. I don't know what does this code chunk do, but I noticed that before the title "<<<  SOLVING SETS OF LINEAR EQUATIONS FOR LINEAR RESPONSE PROPERTY >>>"  this code generates two additional integration that take a long time.

Am I messing up with other property calculations in DIRAC?

Best regards.

Luis