Changeset 979 for trunk/DOC/chantiers


Ignore:
Timestamp:
Jun 5, 2013, 2:41:09 PM (12 years ago)
Author:
emillour
Message:

Common dynamics:
Updates in the dynamics (seq and ) to keep up with updates
in LMDZ5 (up to LMDZ5 trunk, rev 1760):

  • General stuff (essentially to keep up with Earth model):
  • Updated makelmdz_fcm and makelmdz (more control on dimension.h, added option -mem, although it is not usefull for now)
  • Updated build_gcm with more control over fcm
  • Updated create_make_gcm (enable looking for code in subdirectories)
  • bibio:
  • updates (just renaming the files actually...) new_unit.F90 => new_unit_m.F90, pchsp_95.F90 => pchsp_95_m.F90 and pchfe_95.F90 => pchfe_95_m.F90
  • filtrez:
  • mod_fft.F90: use more baseline CPP directives for preprocessor compatibility
  • mod_filtre_fft_loc.F90: added this new file
  • filtreg_mod.f90: added calls to init_..._loc
  • filtreg.F: fixed calls to DGEMM into SGEMM (preprocessing does the switch)
  • dyn3d:
  • removed obsolete files: etat0_netcdf.F90 limit_netcdf.F90

pres2lev.F90

  • added new file : pres2lev_mod.F90 (module containing "old" pres2lev)
  • gcm.F: changed args to call to inidissip (added arg "vert_prof_dissip")
  • inidissip.F90: added arg "vert_prof_dissip" and the "earth model" discterizations (flagged with "planet_type=="earth")
  • comdissnew.h: added 'vert_prof_dissip' to the common block
  • guide_mod.F90: added the "use pres2lev_mod"
  • conf_gcm.F: cosmetics, and evaluation of vert_prof_dissip, (and also of dissip_* factors, for Earth model)
  • comconst.h : added dissip_factz,dissip_zref variables (for Earth mode dissip)
  • dyn3dpar:
  • removed obsolete files: etat0_netcdf.F90 limit_netcdf.F90

pres2lev.F90 mod_const_para.F90

  • added new files: pres2lev_mod.F90 (module containing "old" pres2lev)

mod_const_mpi

  • abort_gcm : better control of abort in parallel mode
  • gcm.F: changed args to call to inidissip (added arg "vert_prof_dissip")
  • inidissip.F90: added arg "vert_prof_dissip" and the "earth model" discterizations (flagged with "planet_type=="earth")
  • comdissnew.h: added 'vert_prof_dissip' to the common block
  • filtreg_p.F : bug correction (array bounds)
  • guide_p_mod.F90 : added the "use pres2lev_mod"
  • conf_gcm.F : cosmetics (and evaluation of vert_prof_dissip , and

also of dissip_* factors, for Earth model)
plus check if "adjust" is indeed not used in OpenMP

  • comconst.h : add dissip_factz,dissip_zref variables (for Earth mode dissip)

EM

Location:
trunk/DOC/chantiers
Files:
2 edited

Legend:

Unmodified
Added
Removed
  • trunk/DOC/chantiers/commit_importants.log

    r841 r979  
    11681168  Now dynamics in seq/MPI/OpenMP/mixed modes yield identical results when
    11691169  running in Newtonian mode (i.e. dynamics only).
     1170
     1171*********************
     1172**** commit_v979 ****
     1173*********************
     1174Ehouarn: Updates in the dynamics (seq and //) to keep up with updates
     1175in LMDZ5 (up to LMDZ5 trunk, rev 1760):
     1176* General stuff (essentially to keep up with Earth model):
     1177- Updated makelmdz_fcm and makelmdz (more control on dimension.h,
     1178  added option -mem, although it is not usefull for now)
     1179- Updated build_gcm with more control over fcm
     1180- Updated create_make_gcm (enable looking for code in subdirectories)
     1181
     1182* bibio:
     1183- updates (just renaming the files actually...) new_unit.F90 => new_unit_m.F90,
     1184  pchsp_95.F90 => pchsp_95_m.F90 and pchfe_95.F90 => pchfe_95_m.F90
     1185
     1186* filtrez:
     1187- mod_fft.F90: use more baseline CPP directives for preprocessor compatibility
     1188- mod_filtre_fft_loc.F90: added this new file
     1189- filtreg_mod.f90: added calls to init_..._loc
     1190- filtreg.F: fixed calls to DGEMM into SGEMM (preprocessing does the switch)
     1191
     1192* dyn3d:
     1193- removed obsolete files: etat0_netcdf.F90 limit_netcdf.F90
     1194                          pres2lev.F90
     1195- added new file : pres2lev_mod.F90 (module containing "old" pres2lev)
     1196- gcm.F: changed args to call to inidissip (added arg "vert_prof_dissip")
     1197- inidissip.F90: added arg "vert_prof_dissip" and the "earth model"
     1198  discterizations (flagged with "planet_type=="earth")
     1199- comdissnew.h: added 'vert_prof_dissip' to the common block
     1200- guide_mod.F90: added the "use pres2lev_mod"
     1201- conf_gcm.F: cosmetics, and evaluation of vert_prof_dissip,
     1202  (and also of dissip_* factors, for Earth model)
     1203- comconst.h : added dissip_factz,dissip_zref variables (for Earth mode dissip)
     1204
     1205* dyn3dpar:
     1206- removed obsolete files: etat0_netcdf.F90 limit_netcdf.F90
     1207                          pres2lev.F90 mod_const_para.F90
     1208- added new files: pres2lev_mod.F90 (module containing "old" pres2lev)
     1209                   mod_const_mpi
     1210- abort_gcm : better control of abort in parallel mode
     1211- gcm.F: changed args to call to inidissip (added arg "vert_prof_dissip")
     1212- inidissip.F90: added arg "vert_prof_dissip" and the "earth model"
     1213  discterizations (flagged with "planet_type=="earth")
     1214- comdissnew.h: added 'vert_prof_dissip' to the common block
     1215- filtreg_p.F : bug correction (array bounds)
     1216- guide_p_mod.F90 : added the "use pres2lev_mod"
     1217- conf_gcm.F : cosmetics (and evaluation of vert_prof_dissip , and
     1218               also of dissip_* factors, for Earth model)
     1219               plus check if "adjust" is indeed not used in OpenMP
     1220- comconst.h : add dissip_factz,dissip_zref variables (for Earth mode dissip)
  • trunk/DOC/chantiers/meschantiers-Ehouarn.txt

    r895 r979  
    1414 
    1515- dernière mise à jour des dynamiques (par rapport au trunk de LMDZ5
    16   "Terrestre" rev. 1678) avec la rev 841.
     16  "Terrestre" rev. 1760) avec la rev 979.
    1717  Reste à faire l'exercice de validation des modifs en //...
    1818  Avec rev 841 et 843 Cp(T) en // OK
    1919  Reste toutefois à paufiner les choses au niveau de calfis_p et des
    2020  domaines d'utilisation de tpot2t_p et t2tpot_p (toutes les taches MPI
    21   font l'intégralité du travail!)
     21  font l'intégralité du travail!). OK.
    2222
    2323- Faire qu'on puisse compiler un exécutable qui serait dans la physique
Note: See TracChangeset for help on using the changeset viewer.