Changeset 1572 for trunk/LMDZ.COMMON/libf/misc

All GCMs:
Further adaptations to keep up with changes in LMDZ5 concerning
physics/dynamics separation (up to rev r2500 of LMDZ5)

• arch:
• remove ifort debug option '-check all', replace it with '-check bounds,format,output_conversion,pointers,uninit' (i.e. get it to stop complaining about copying into temporary arrays)

• dyn3d_common:
• comconst_mod.F90 : add ngroup

• dyn3d:
• gcm.F90 : minor bug fix (arguments to a call_abort())
• leapfrog.F90 : recompute geopotential for bilan_dyn outputs
• conf_gcm.F90 : read "ngroup" from run.def
• groupe.F , groupeun.F : ngroup no longer a local parameter

• dyn3d_par:
• conf_gcm.F90 : read "ngroup" from run.def
• groupe_p.F , groupeun_p.F : ngroup no longer a local parameter

• misc:
• regr1_step_av_m.F90 : removed (not used)

• phy_common:
• mod_phys_lmdz_mpi_transfert.F90 , mod_phys_lmdz_mpi_data.F90 : change is_north_pole and is_south_pole to is_north_pole_dyn and is_south_pole_dyn
• mod_phys_lmdz_omp_data.F90 : introduce is_nort_pole_phy and is_south_pole_phy

• dynphy_lonlat:
• mod_interface_dyn_phys.F90 : use is_north_pole_dyn and is_south_pole_dyn
• calfis_p.F : use is_north_pole_dyn and is_south_pole_dyn

• phyvenus:
• physiq_mod , write_hist*.h : use is_north_pole_phy and is_south_pole_phy to correctly compute mesh area at poles to send to hist*nc files.

• phytitan:
• physiq_mod , write_hist*.h : use is_north_pole_phy and is_south_pole_phy to correctly compute mesh area at poles to send to hist*nc files.

EM