4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
5 |
============================== |
============================== |
6 |
|
|
7 |
|
o eesupp & mdsio (SingleCpuIO): |
8 |
|
- move mapping to global io-buffer inside gather_2d/scater_2d: this save |
9 |
|
memory (1 less 2D global RL array); skip 1 buffer copy; and only send/ |
10 |
|
receive real*4 array (instead of real*8) to read/write 32.bit files. |
11 |
|
(new template for gather_2d/scater_2d, new set of S/R MDS_PASS_R4,8toRL,S |
12 |
|
and remove new mapping S/R) |
13 |
|
o pkg/mdsio: in preparation for topology-generator within pkg/exch2: |
14 |
|
- Change declaration of SingleCpuIO buffer, reduced to only 1 dim.: |
15 |
|
xyBuffer_size, which is set to W2_ioBufferSize from W2_EXCH2_PARAMS.h |
16 |
|
when using exch2. Add S/R to map global io-buffer to global model array. |
17 |
|
- uses exch2_global_Nx,y instead of exch2_domain_nxt,nyt. |
18 |
|
|
19 |
|
checkpoint61m |
20 |
|
o write_utils : change WRITE_1D_R8 to WRITE_1D_RL |
21 |
|
and add 3 other S/R for other types (R4,R8,RS). |
22 |
|
Change calls to WRITE_0D/1D_R[4,8,S,L] to match the type of argument. |
23 |
|
o eesupp/inc/CPP_EEMACROS.h (EXCH & GLOBAL_SUM/MAX macros): |
24 |
|
- ensure suffix (_R4,_R8,_RS,_RL) correspondance between macro and S/R ; |
25 |
|
- reduce number of macro definitions (keep only those which are used); |
26 |
|
can always call directly the corrresponding S/R if needed. |
27 |
|
- everywhere in the code: change macros (EXCH & GLOBAL_SUM/MAX) suffix |
28 |
|
_R4/_R8 to _RS/_RL when applied to _RS/_RL variable. |
29 |
|
Fixed wrong type of arg in: |
30 |
|
atm2d/put_ocnvars.F cheapaml/cheapaml.F ocn_compon_interf/ocn_apply_import.F |
31 |
|
ctrl/ctrl_map_ini_ecco.F seaice/seaice_do_diags.F seaice/seaice_model.F |
32 |
|
To fix: global_sum of var in common block is wrong (multi-threaded) in: |
33 |
|
dic/dic_atmos.F ebm/ebm_area_t.F ebm/ebm_zonalmean.F sbo/sbo_calc.F |
34 |
|
To fix: aim_v23/aim_do_co2.F (BARRIER & GLOBAL_SUM within bi,bj loops) |
35 |
|
o adjustment.cs-32x32x1: test "blank tiles" when compiling with mpi; 2nd |
36 |
|
"nlfs" mpi test is going to stop when checking for unconnected tile egdes. |
37 |
|
o eesupp/src: new S/R to stop everyone when 1 (or more) Proc. find an error |
38 |
|
o pkg/exch2: new S/R EXCH2_CHECK_DEPTHS to check that unconnected tile |
39 |
|
edges are closed (zero depth); might happen with "blank tiles"; |
40 |
|
for now, disabled if using OBCS. |
41 |
|
o tools/gemake2: delete modifs from March 27-29 2009 (with option "-mpi": |
42 |
|
link *_mpi files from "MODS" dir) to return to a strait-forward behavior. |
43 |
|
o move read_write.F to write_utils.F ; new parameter (maxLengthPrt1D) |
44 |
|
to reduce length of 1.D array print in STDOUT. |
45 |
|
o adjustment.cs-32x32x1: changed to oceanic test with 1 continent |
46 |
|
+ use 48 tiles (16x8): this will allow to test "blank tiles". |
47 |
|
o start to test MPI + multi-threaded (change testreport to skip tests |
48 |
|
where number of tiles do not match). |
49 |
|
o genmake2: add option "-omp" to turn on Open-MP ; |
50 |
|
passed to testreport (if using -mth) ; merge few (*+mth) optfiles. |
51 |
|
o pkg/exch2 (+ utils/exch2): ignore blank tiles (instead of removing them) |
52 |
|
and keep the original tile numbering. Note: compatible with old |
53 |
|
W2_EXCH2_TOPOLOGY.h & w2_e2setup.F topology files. |
54 |
|
|
55 |
|
checkpoint61l |
56 |
|
o Overlaps had been forgotten in calculating ijk keys : fixed |
57 |
|
o fix cyrus makedepend for gcc 4.4 |
58 |
|
o experiment fizhi-cs-aqualev20: |
59 |
|
switching to new version of S/R GETPWHERE (#define TRY_NEW_GETPWHERE) |
60 |
|
and generate new output. Note: new output is much closer to the old |
61 |
|
one (v1.8) generated with pgf77 -O2 on Mar 14 2007 than it is from |
62 |
|
the previous one (v1.9) generated with pgf77 -O0 on Jun 13 2007. |
63 |
|
o pkg/fizhi: |
64 |
|
- change to avoid indices going deliberately over array-bounds |
65 |
|
- remove unnecessary MPI stuff |
66 |
|
- perpetual spring equinox with "#define FIZHI_USE_FIXED_DAY" |
67 |
|
- new version of S/R GETPWHERE available with #define TRY_NEW_GETPWHERE: |
68 |
|
does not go over array-bounds and therefore returns reliable output. |
69 |
|
- in fizhi_driver.F & fizhi_turb.F, fix "get_alarm" calls: |
70 |
|
this S/R has 4 output arguments, which are used inside "get_alarm"; |
71 |
|
when it is called with the same variable "ndum" in place of several |
72 |
|
output arguments, this produces weird side effect inside "get_alarm" |
73 |
|
(e.g., datein=dates(n) is overwritten by timein=times(n) since both |
74 |
|
datein & timein, resp. 2nd and 3rd arg., are pointed to the same "ndum" |
75 |
|
address) and produces error. |
76 |
|
o tools/gemake2: (March 27-29 2009) |
77 |
|
with option "-mpi": link *_mpi files from "MODS" dir (this is what |
78 |
|
testreport does). |
79 |
|
o lab_sea experiment: change from 2x1 tiles to 2x2 tiles and regenerate output |
80 |
|
o pkg/seaice: fix bug in new version of S/R SEAICE_LSR.F: |
81 |
|
update lab_sea.lsr & lab_sea.salt_plume results |
82 |
|
o pkg/seaice: |
83 |
|
Major overhaul of the LSOR solver: change discretization from original |
84 |
|
finite differences to finite volume. The old routine is still in the |
85 |
|
file seaice_lsr.F and can be turned on with the CPP-flag |
86 |
|
SEAICE_OLD_AND_BAD_DISCRETIZATION |
87 |
|
Along with this change, the discretization of the strain rates is changed. |
88 |
|
The same flag turns on the old discretization in seaice_calc_strainrates.F |
89 |
|
All verification experiments with seaice are affected (except for |
90 |
|
seaice_exf_offline); global_ocean.cs32x15/input.icedyn recovers the old |
91 |
|
results by setting the above flag. |
92 |
|
Further changes involve changes in defaults: |
93 |
|
- ALLOW_FLOODING is defined and turned on by default |
94 |
|
- SEAICEadvSnow = .true. is now the default |
95 |
|
- SEAICEadvSalt = .true. is now the default |
96 |
|
- SEAICEadvAge = .true. is now the default |
97 |
|
- SEAICE_clipVelocities = .false. is now the default |
98 |
|
(as per J. Zhang's recommendation) |
99 |
|
- B-grid, and thus not tested: SEAICE_TEST_ICE_STRESS_1/EXPLICIT_SSH_SLOPE |
100 |
|
is defined, SEAICE_TEST_ICE_STRESS_1 is renamed into SEAICE_BICE_STRESS |
101 |
|
- seaice_growth: replace computation of UG by a simple copy from wspeed |
102 |
|
|
103 |
|
checkpoint61k |
104 |
|
o pkg/seaice: a few bug fixes mostly concerning the metric terms: |
105 |
|
seaice_lsr: fixed a few metric terms, |
106 |
|
affects lab_sea.lsr, lab_sea.salt_plume, seaice_obcs |
107 |
|
add more debugging diagnostics |
108 |
|
seaice_evp: removed superfluous many metric terms of FV discretization |
109 |
|
of stress divergence |
110 |
|
affects lab_sea, lab_sea.hb87 |
111 |
|
seaice_ocean_stress: removed superfluous many metric terms of FV discretization |
112 |
|
of stress divergence for non-evp solver |
113 |
|
(this part is not tested anywhere) |
114 |
|
seaice_calc_strainrates: fixed a bug in no-slip boundary conditions, |
115 |
|
they were effectively only half slip |
116 |
|
affects lab_sea.hb87 |
117 |
|
updated all relevant verification experiments (except lab_sea//output_adm.txt |
118 |
|
which was left over, but fails at level 7 & 6) |
119 |
|
|
120 |
checkpoint61j |
checkpoint61j |
121 |
o pkg/seaice: turn a few hard wired parameters into run time parameters |
o pkg/seaice: turn a few hard wired parameters into run time parameters |
122 |
o pkg/seaice/exf/monitor: turn netcdf coordinate variable T into model |
o pkg/seaice/exf/monitor: turn netcdf coordinate variable T into model |