| 4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 5 |
============================== |
============================== |
| 6 |
|
|
| 7 |
|
o model/src: |
| 8 |
|
- improve vectorization also for adjoint code of |
| 9 |
|
solve_tridiagonal and solve_pentadiagonal. As this requires up to |
| 10 |
|
5 new 3D-fields, this code is only turned on for the combination of |
| 11 |
|
ALLOW_AUTODIFF_TAMC and TARGET_NEC_SX. |
| 12 |
|
- remove hard-wiring of CPP-flag ALLOW_SOLVERS_KLOOPINSIDE to |
| 13 |
|
ALLOW_AUTODIFF. This flag now needs to be set outside of |
| 14 |
|
solve_tri/pentadiagonal.F |
| 15 |
|
o pkg/exf: |
| 16 |
|
- fix previous modif in exf_readparms.F (revision 1.66: treatment of |
| 17 |
|
optional namelist EXF_NML_OBCS). |
| 18 |
|
- add many debug-print (to trace down where pkg/cal error msg comes from) |
| 19 |
|
o pkg/mom_vecinv/mom_vi_hdissip.F |
| 20 |
|
- rearrange loops so that vectorization is improved and in particular |
| 21 |
|
the adjoint routine vectorizes. The code is a little longer now, but I |
| 22 |
|
think easier to read. |
| 23 |
|
Note: CPP-flag MOM_VI_ORIGINAL_VISCA4 is not tested. |
| 24 |
|
o pkg/seaice: |
| 25 |
|
- allow to specify initial seaice velocity from files (uIceFile,vIceFile) |
| 26 |
|
o pkg/exch2: |
| 27 |
|
- alternative code (without tile x tile matrix) for cumulated (global) sum; |
| 28 |
|
used as default; keep using matrix version in adjustment.cs-32x32x1. |
| 29 |
|
o tools/build_options: |
| 30 |
|
- after merging some MPI and/or multi-threads specific optfiles so that |
| 31 |
|
the modified default optfile support both built (using genmake2 -mpi |
| 32 |
|
for the MPI built, using genmake2 -omp for OpenMP built), |
| 33 |
|
move now obsolete optfiles to tools/build_options/unsupported/ |
| 34 |
|
(add copy of email sent to support list in doc/optfiles_changes.txt) |
| 35 |
|
|
| 36 |
|
checkpoint63c (2011/09/28) |
| 37 |
|
o pkg/ebm: |
| 38 |
|
- fix ebm_zonalmean.F for several procs in Ydir (nPy > 1); update output of |
| 39 |
|
test exp. global_ocean_ebm (affects machine truncation, 12 digits for cg2d) |
| 40 |
|
o pkg/seaice (SItracer pickup) |
| 41 |
|
* Simple code for pickup of SItracer fields (#define ALLOW_SITRACER) |
| 42 |
|
To be done: |
| 43 |
|
- thorough testing of missing fields in S/R seaice_read_pickup |
| 44 |
|
- thorough matching between generic tracers and specific fields used |
| 45 |
|
(e.g. volume-weighted age, area-weighted age, salinity, ...) |
| 46 |
|
o pkg/grdchk: |
| 47 |
|
- all procs calculate and print finite-difference gradient (for testreport) |
| 48 |
|
- change description output used by testreport; |
| 49 |
|
- AD testing: also compare and report Finite-difference gradient; |
| 50 |
|
=> update all output_adm*.txt |
| 51 |
|
o test exp. global_ocean_ebm: |
| 52 |
|
- switch back useRealFreshWaterFlux (no problem anymore with run-off); |
| 53 |
|
and turn on exactConserve; update (fwd) output. |
| 54 |
|
o pkg/seaice (LSR): |
| 55 |
|
- print to STDOUT number of solver iterations at SEAICE_monFreq interval. |
| 56 |
|
- modify solver to deal cleanly with u/v switch at face edges (CS-grid): |
| 57 |
|
* identical results with non-CS topology (but save up to half of EXCH calls) |
| 58 |
|
* no much changes in cs-32 verification tests since no ice at face edges |
| 59 |
|
(differences due to convergence criteria, since nb of iterations changed) |
| 60 |
|
* expect some differences on lat-lon-cap grid. |
| 61 |
|
update FWD results of global_ocean.cs32x15 test exp. (icedyn & seaice) |
| 62 |
|
o pkg/exch2: |
| 63 |
|
- add new CPP option to disable Cumul-Sum matrix (to save storage when using |
| 64 |
|
many tiles); until alternative coding is added, just add a stop. |
| 65 |
|
o pkg/kpp: |
| 66 |
|
- move ptracer/salt correction to KPP diffusivity from calc_3d_diffusivity.F |
| 67 |
|
to S/R KPP_CALC_DIFF_PTR (+ add tracer-index to argument list). |
| 68 |
|
o optfiles: |
| 69 |
|
- merged few MPI specific (*+mpi*) and multi-threads specific (*+mth) optfiles |
| 70 |
|
into corresponding standard one, which now support different build: |
| 71 |
|
(when using "genmake2 -mpi" for MPI built, "genmake2 -omp" for OpenMP built) |
| 72 |
|
- add "darwin_amd64_gfortran" optfile (initially from Constantinos, tested by Gael) |
| 73 |
|
o pkg/obcs (seaice): |
| 74 |
|
- add missing EXCH for uIce,vIce after applying OBCS; |
| 75 |
|
- remove S/R EXF_SWAPFFIELDS_X/Y , disable S/R EXF_SET_OBCS_X/Y by adding |
| 76 |
|
argument = number of levels to S/R EXF_SWAPFFIELDS_XZ/YZ and S/R |
| 77 |
|
EXF_SET_OBCS_XZ/YZ ; update obcs_exf_load.F; |
| 78 |
|
- fix EXF_SET_OBCS_XZ/YZ for multi-threaded (assuming loaded arrays are shared) |
| 79 |
|
- verification exp. seaice_obcs: split domain in 2 tiles ; add SIZE.h_mpi |
| 80 |
|
and eedata.mth for MPI and mutlti-threaded testing. |
| 81 |
|
o pkg/flt: |
| 82 |
|
- make FLT EXCH working with multi-threads (and MPI + multi-threads); |
| 83 |
|
- use buffer for floats IO (trajectories & profiles); size set in FLT_SIZE.h; |
| 84 |
|
- fix pickup and trajectories + profiles output with multi-threads; |
| 85 |
|
- add parameter to select group of fields to write along trajectory and |
| 86 |
|
profile (default unchanged); |
| 87 |
|
- only call PORT_RAND* function when flt_noise <> 0; |
| 88 |
|
- add multi-threaded test for verification exp. flt_example. |
| 89 |
|
o eesupp/src: |
| 90 |
|
- fix S/R ALL_PROC_DIE for MPI+multi-threads. |
| 91 |
|
- USE_OMP_THREADING: check that number of threads (from eedata) matches |
| 92 |
|
actual number (from env.var. setting). |
| 93 |
o pkg/ebm: |
o pkg/ebm: |
| 94 |
- fix multi-threading in ebm_area_t.F ; clean-up EBM.h header file. |
- fix multi-threading in ebm_area_t.F ; clean-up EBM.h header file. |
| 95 |
- add multi-threaded test for verification exp. global_ocean_ebm |
- add multi-threaded test for verification exp. global_ocean_ebm |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch