Subprogram calc_psi0_corr (sigmat, chit, u, mu, isp)

This function is not consistent with compute_psi0_corr, though I don't think the values are ever used - ALS

Subprogram diagnostics_mod::initialize_diagnostics

Not clear why this should only happen on me_world==0. Do the others not use these diagnostics? - ALS

Subprogram equil_mod::back_vel_mod (sigmat, chi, vexb2)

LV: back_vel_mod is called from several places, many of which irrespective of nl_strong But c_equ(48,:,:) is computed only if species(1)nl_strong Before the logic of the code is fixed, a patch is applied here

Subprogram equil_mod::back_vel_tor (sigmat, chi, ue_phi)

LV: back_vel_tor is called from several places, many of which irrespective of nl_strong But c_equ(47,:,:) is computed only if species(1)nl_strong Before the logic of the code is fixed, a patch is applied here

Subprogram equil_mod::calculate_other_normalizations

zvvt2, zpvt2, and zpkkt2 should probably be set to 0 at the beginning of this loop, not the i loop - ALS

Subprogram fhs_build_mat (pic1_loc, pic2_loc, npart_loc, isp)

No support for OpenMP for the moment

Subprogram load_alp (pic2_loc, work_rev_loc, isp)

It appears "alp" is never loaded!! ALS

Subprogram part::build_vpaob

ALS Should psi_max, psi_min, really depend on the last species in the loop?

Subprogram part::diagnose_particles

Ions are suposed to be the first species which is not always the case.

Electrons are suposed to be the second species which is not always the case.

Subprogram part::get_properties (ip, isp, sigmat, chit, en, b, psi0can, f0, zf0, f0_max, T, en_hamilt)

Is it correct to add zecor to ze since it will be also added in vf0 for example ?

Subprogram part::get_rhs (ip, isp, pvol, vdr, vdchi, vgc, vin, vpval, genelec, dvpadt0, dvpadt1, gyroapar, ze, zecor, dH0_dt_correction, pot_equil, pot_equil0, distribution, ut, mut, b, zvth2, nval, Epsi, psi0t, ngrad, tgrad, vpgrad, dfvpllb_dpsi, dfvpllb_den, rhs, dpotdpsi, omegasq_corr, df0_dpsi_prof, df0_dchi_prof, dpsi_prof_dt, dchi_prof_dt, rhs_type, mgradphi, vgb, vpressure, mgradapar, hh_sympl, chit, dchi_dt1, iter, ffinit, dPsidt0)

epsilon_hat is set to 0.0 here to avoid uninitialized value for case 1 and 3. For those latter cases it should be checked whether epsilon_hat should be computed or not. Note that until now, we relied on the compiler to initialize epsilon_hat to 0.0.

Subprogram part::helical_flux_label (sigmat, xiang, irloc, dr_flux_inv, ipol)

func_torpsi takes psi as argument but it is defined for s (see def. in equil.F90).

Subprogram part::push (pic1_loc, pic2_loc, pic3_loc, work1_loc, npart_loc, isp, iter)

Why is this different in the merged verion? IPP/McMillan merge 16/12/2014

Subprogram read_attributes

Case NSEL_TYPE_4 used to raise this error, but I removed it since the linear benchmarks were using
nsel_profile = NSEL_TYPE_4. Potential problem? - ALS