4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
5 |
============================== |
============================== |
6 |
|
|
7 |
|
o pkg/layers: |
8 |
|
- Include potential density as new coordinate (Thanks to David Munday) |
9 |
|
o pkg/dic: |
10 |
|
- Remove subroutine CALC_PCO2_APPROX_CO3 from carbon_chem.F and add |
11 |
|
carbonate computation/output to CALC_PCO2_APPROX |
12 |
|
o model/src: |
13 |
|
- improve vectorization also for adjoint code of |
14 |
|
solve_tridiagonal and solve_pentadiagonal. As this requires up to |
15 |
|
5 new 3D-fields, this code is only turned on for the combination of |
16 |
|
ALLOW_AUTODIFF_TAMC and TARGET_NEC_SX. |
17 |
|
- remove hard-wiring of CPP-flag ALLOW_SOLVERS_KLOOPINSIDE to |
18 |
|
ALLOW_AUTODIFF. This flag now needs to be set outside of |
19 |
|
solve_tri/pentadiagonal.F |
20 |
|
o pkg/exf: |
21 |
|
- fix previous modif in exf_readparms.F (revision 1.66: treatment of |
22 |
|
optional namelist EXF_NML_OBCS). |
23 |
|
- add many debug-print (to trace down where pkg/cal error msg comes from) |
24 |
|
o pkg/mom_vecinv/mom_vi_hdissip.F |
25 |
|
- rearrange loops so that vectorization is improved and in particular |
26 |
|
the adjoint routine vectorizes. The code is a little longer now, but I |
27 |
|
think easier to read. |
28 |
|
Note: CPP-flag MOM_VI_ORIGINAL_VISCA4 is not tested. |
29 |
|
o pkg/seaice: |
30 |
|
- allow to specify initial seaice velocity from files (uIceFile,vIceFile) |
31 |
|
o pkg/exch2: |
32 |
|
- alternative code (without tile x tile matrix) for cumulated (global) sum; |
33 |
|
used as default; keep using matrix version in adjustment.cs-32x32x1. |
34 |
|
o tools/build_options: |
35 |
|
- after merging some MPI and/or multi-threads specific optfiles so that |
36 |
|
the modified default optfile support both built (using genmake2 -mpi |
37 |
|
for the MPI built, using genmake2 -omp for OpenMP built), |
38 |
|
move now obsolete optfiles to tools/build_options/unsupported/ |
39 |
|
(add copy of email sent to support list in doc/optfiles_changes.txt) |
40 |
|
|
41 |
|
checkpoint63c (2011/09/28) |
42 |
|
o pkg/ebm: |
43 |
|
- fix ebm_zonalmean.F for several procs in Ydir (nPy > 1); update output of |
44 |
|
test exp. global_ocean_ebm (affects machine truncation, 12 digits for cg2d) |
45 |
|
o pkg/seaice (SItracer pickup) |
46 |
|
* Simple code for pickup of SItracer fields (#define ALLOW_SITRACER) |
47 |
|
To be done: |
48 |
|
- thorough testing of missing fields in S/R seaice_read_pickup |
49 |
|
- thorough matching between generic tracers and specific fields used |
50 |
|
(e.g. volume-weighted age, area-weighted age, salinity, ...) |
51 |
|
o pkg/grdchk: |
52 |
|
- all procs calculate and print finite-difference gradient (for testreport) |
53 |
|
- change description output used by testreport; |
54 |
|
- AD testing: also compare and report Finite-difference gradient; |
55 |
|
=> update all output_adm*.txt |
56 |
|
o test exp. global_ocean_ebm: |
57 |
|
- switch back useRealFreshWaterFlux (no problem anymore with run-off); |
58 |
|
and turn on exactConserve; update (fwd) output. |
59 |
|
o pkg/seaice (LSR): |
60 |
|
- print to STDOUT number of solver iterations at SEAICE_monFreq interval. |
61 |
|
- modify solver to deal cleanly with u/v switch at face edges (CS-grid): |
62 |
|
* identical results with non-CS topology (but save up to half of EXCH calls) |
63 |
|
* no much changes in cs-32 verification tests since no ice at face edges |
64 |
|
(differences due to convergence criteria, since nb of iterations changed) |
65 |
|
* expect some differences on lat-lon-cap grid. |
66 |
|
update FWD results of global_ocean.cs32x15 test exp. (icedyn & seaice) |
67 |
|
o pkg/exch2: |
68 |
|
- add new CPP option to disable Cumul-Sum matrix (to save storage when using |
69 |
|
many tiles); until alternative coding is added, just add a stop. |
70 |
|
o pkg/kpp: |
71 |
|
- move ptracer/salt correction to KPP diffusivity from calc_3d_diffusivity.F |
72 |
|
to S/R KPP_CALC_DIFF_PTR (+ add tracer-index to argument list). |
73 |
|
o optfiles: |
74 |
|
- merged few MPI specific (*+mpi*) and multi-threads specific (*+mth) optfiles |
75 |
|
into corresponding standard one, which now support different build: |
76 |
|
(when using "genmake2 -mpi" for MPI built, "genmake2 -omp" for OpenMP built) |
77 |
|
- add "darwin_amd64_gfortran" optfile (initially from Constantinos, tested by Gael) |
78 |
|
o pkg/obcs (seaice): |
79 |
|
- add missing EXCH for uIce,vIce after applying OBCS; |
80 |
|
- remove S/R EXF_SWAPFFIELDS_X/Y , disable S/R EXF_SET_OBCS_X/Y by adding |
81 |
|
argument = number of levels to S/R EXF_SWAPFFIELDS_XZ/YZ and S/R |
82 |
|
EXF_SET_OBCS_XZ/YZ ; update obcs_exf_load.F; |
83 |
|
- fix EXF_SET_OBCS_XZ/YZ for multi-threaded (assuming loaded arrays are shared) |
84 |
|
- verification exp. seaice_obcs: split domain in 2 tiles ; add SIZE.h_mpi |
85 |
|
and eedata.mth for MPI and mutlti-threaded testing. |
86 |
|
o pkg/flt: |
87 |
|
- make FLT EXCH working with multi-threads (and MPI + multi-threads); |
88 |
|
- use buffer for floats IO (trajectories & profiles); size set in FLT_SIZE.h; |
89 |
|
- fix pickup and trajectories + profiles output with multi-threads; |
90 |
|
- add parameter to select group of fields to write along trajectory and |
91 |
|
profile (default unchanged); |
92 |
|
- only call PORT_RAND* function when flt_noise <> 0; |
93 |
|
- add multi-threaded test for verification exp. flt_example. |
94 |
|
o eesupp/src: |
95 |
|
- fix S/R ALL_PROC_DIE for MPI+multi-threads. |
96 |
|
- USE_OMP_THREADING: check that number of threads (from eedata) matches |
97 |
|
actual number (from env.var. setting). |
98 |
|
o pkg/ebm: |
99 |
|
- fix multi-threading in ebm_area_t.F ; clean-up EBM.h header file. |
100 |
|
- add multi-threaded test for verification exp. global_ocean_ebm |
101 |
|
|
102 |
|
checkpoint63b (2011/08/29) |
103 |
|
o pkg/ebm: |
104 |
|
- fix sign of runoff; |
105 |
|
- update directly forcing fields (Qnet,EmPmR) to use standard |
106 |
|
external_forcing_surf.F (remove ebm_forcing_surf.F); |
107 |
|
- avoid loading forcing fields a 2nd time (1rst time from ini_forcing.F); |
108 |
|
- set either tauThetaClimRelax (in data) or tauThetaZonRelax (in data.ebm) |
109 |
|
and apply relaxation towards corresponding SST field (2-D map or ZonAver). |
110 |
|
- update global_ocean_ebm parameters file data and update output (fwd+ad). |
111 |
|
o verification/matrix_example: |
112 |
|
- split domain in 8 tiles (2x4); add SIZE.h_mpi and eedata.mth for |
113 |
|
MPI and mutlti-threaded testing. |
114 |
|
o pkg/matrix: |
115 |
|
- change name of S/R MATRIX_INIT to S/R MATRIX_INIT_VARIA ; |
116 |
|
- fix for multi-tiles (add bi,bj indices to exp/impMatrixCounter); |
117 |
|
- fix for multi-threaded (remove DATA statement); |
118 |
|
- use standard DIFFERENT_MULTIPLE function; |
119 |
|
- add MATRIX_OPTIONS.h |
120 |
|
o verification/matrix_example: |
121 |
|
To allow to split the domain in several tiles, change the domain size |
122 |
|
from 31x31 to 32x32 by adding 1 row (on Southern side) + 1 column (on |
123 |
|
Western side) of land points. |
124 |
|
o (*_)write_pickup: |
125 |
|
- fix few bugs in mypackage_read/write_pickup.F |
126 |
|
- in MDS_WR_METAFILES calls, replace "myTime" by a local copy into |
127 |
|
array of length 1 (was caught by compiler syntax-check). |
128 |
|
o pkg/diagnostics (velocity potential solver): |
129 |
|
- fix initialisation of main diagonal matrix (for overlap values) |
130 |
|
o tools/build_options: |
131 |
|
- towards single optfile for MPI and non-MPI built: merge few +mpi optfile |
132 |
|
into previously non MPI one (linux_am64 pgf77,open64,pathf90,ifort11) |
133 |
|
- linux_amd64_gfortran: merge +mpi_generic version into standard optfile: |
134 |
|
* move IEEE flags to DEVEL section; take IEEE flags from +mpi_generic ; |
135 |
|
* FAST case: for now, keep -O3 with ini_masks_etc.F in NOOPTFILES list; |
136 |
|
and use "-devel" testreport option for standard testing on baudelaire. |
137 |
|
o pkg/grdchk: |
138 |
|
- Assume that all gradient-check position are within the same tile |
139 |
|
corresponding to global indices (iGloTile,jGloTile) and that |
140 |
|
positions (i,j indices or nbeg) are local to this tile; |
141 |
|
- set tile local-indices and proc. number according to tile global-indices; |
142 |
|
- remove grdchkwhichproc from namelist (retired param); |
143 |
|
- send adjoint gradient to all procs so that all procs print it to STDOUT; |
144 |
|
- switch on MPI test for global_ocean.90x40x15 AD test experiment. |
145 |
|
o adding package bbl (Bottom Boundary Layer) |
146 |
|
description in MITgcm/pkg/bbl/bbl_description.tex |
147 |
|
example/test experiment in MITgcm_contrib/bbl |
148 |
|
o adjoint verifications |
149 |
|
- changed 5 experiments to multi-tiles |
150 |
|
* global_ocean.90x40x15 |
151 |
|
* obcs_ctrl |
152 |
|
* offline_exf_seaice |
153 |
|
* tutorial_dic_adjoffline |
154 |
|
* tutorial_tracer_adjsens |
155 |
|
changed reference output for those that crossed tile boundaries |
156 |
|
(changed choices of gradient check positions) |
157 |
|
- with same tiling, switch on MPI test for the last 4 (add SIZE.h_mpi) |
158 |
|
- temporary bug fix in grdchk_loc.F for obcs |
159 |
|
o genmake2 & testreport |
160 |
|
- add target "obj" (just for OBJFILES files) to Makefile; |
161 |
|
- add option to specify FEXTRAFLAGS on command line. |
162 |
o pkg/mypackage: |
o pkg/mypackage: |
163 |
- remove bi,bj arguments from argument list of S/R MYPACKAGE_CALC_RHS |
- remove bi,bj arguments from argument list of S/R MYPACKAGE_CALC_RHS |
164 |
|
|