| 4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 5 |
============================== |
============================== |
| 6 |
|
|
| 7 |
o tools/gemake2: delete modifs from March 27-29 2009 (with option "-mpi": |
o SingleCpuIO: |
| 8 |
|
fix bug put before chkp61n (May 11), when more than 2 MPI cpu are used. |
| 9 |
|
o pkg/seaice: |
| 10 |
|
put exchanged variable in (local) common block for MPI+Multi-Threads |
| 11 |
|
to work (necessary with exch1, and for now with exch2 too). |
| 12 |
|
o verification: to get more experiments tested with mpi, mth and mpi+mth, |
| 13 |
|
change tiling of: |
| 14 |
|
aim.5l_Equatorial_Channel, aim.5l_LatLon, dome, exp4, global_ocean.90x40x15, |
| 15 |
|
hs94.cs-32x32x5, MLAdjust, offline_exf_seaice & tutorial_plume_on_slope |
| 16 |
|
update output of exp: |
| 17 |
|
exp4, global_ocean.90x40x15(+.dwslp), MLAdjust(all 5), |
| 18 |
|
tutorial_plume_on_slope and hs94.cs-32x32x5 (use default rotationPeriod) |
| 19 |
|
+ ideal_2D_ocean: use default rotation period + test Global_Sum_SINGLECPU |
| 20 |
|
(note: output.txt generated on faulks using only 1 tile) |
| 21 |
|
o eesupp & mdsio (SingleCpuIO): |
| 22 |
|
- remove MISSING_TILE_IO (from MDSIO_OPTIONS.h), replaced by run-time |
| 23 |
|
parameter "W2_useE2ioLayOut" (read from file "data.exch2"). |
| 24 |
|
Note: was not effective for globalFile & useSingleCpuIO=F; now fixed. |
| 25 |
|
- move MDSIO_SCPU.h (pkg/mdsio) to EEBUFF_SCPU.h (eesupp/inc). |
| 26 |
|
- fix GLOBAL_SUM_SINGLECPU when using Exch2; re-use same buffers and same |
| 27 |
|
gather/scatter S/R as with SingleCpuIO (=> 1 less 2D global RL array). |
| 28 |
|
o pkg/exch2: |
| 29 |
|
- new code to set-up W2-Exch2 topology (replace matlab-topology-generator): |
| 30 |
|
read parameter file "data.exch2" if it exists; otherwise try default regular |
| 31 |
|
cube without blank-tile; new header file "W2_EXCH2_SIZE.h"; |
| 32 |
|
- update matlab-topology-generator (exch2_setup_cs6_print.m) to be compatible |
| 33 |
|
with new code (not to be used, just as a backup solution). |
| 34 |
|
- add examples (e.g., for lat-lon-cap_120) of parameter file "data.exch2" |
| 35 |
|
in utils/exch2/input. |
| 36 |
|
|
| 37 |
|
checkpoint61n |
| 38 |
|
o eesupp & mdsio (SingleCpuIO): |
| 39 |
|
- move mapping to global io-buffer inside gather_2d/scater_2d: this save |
| 40 |
|
memory (1 less 2D global RL array); skip 1 buffer copy; and only send/ |
| 41 |
|
receive real*4 array (instead of real*8) to read/write 32.bit files. |
| 42 |
|
(new template for gather_2d/scater_2d, new set of S/R MDS_PASS_R4,8toRL,S |
| 43 |
|
and remove new mapping S/R) |
| 44 |
|
o pkg/mdsio: in preparation for topology-generator within pkg/exch2: |
| 45 |
|
- Change declaration of SingleCpuIO buffer, reduced to only 1 dim.: |
| 46 |
|
xyBuffer_size, which is set to W2_ioBufferSize from W2_EXCH2_PARAMS.h |
| 47 |
|
when using exch2. Add S/R to map global io-buffer to global model array. |
| 48 |
|
- uses exch2_global_Nx,y instead of exch2_domain_nxt,nyt. |
| 49 |
|
|
| 50 |
|
checkpoint61m |
| 51 |
|
o write_utils : change WRITE_1D_R8 to WRITE_1D_RL |
| 52 |
|
and add 3 other S/R for other types (R4,R8,RS). |
| 53 |
|
Change calls to WRITE_0D/1D_R[4,8,S,L] to match the type of argument. |
| 54 |
|
o eesupp/inc/CPP_EEMACROS.h (EXCH & GLOBAL_SUM/MAX macros): |
| 55 |
|
- ensure suffix (_R4,_R8,_RS,_RL) correspondance between macro and S/R ; |
| 56 |
|
- reduce number of macro definitions (keep only those which are used); |
| 57 |
|
can always call directly the corrresponding S/R if needed. |
| 58 |
|
- everywhere in the code: change macros (EXCH & GLOBAL_SUM/MAX) suffix |
| 59 |
|
_R4/_R8 to _RS/_RL when applied to _RS/_RL variable. |
| 60 |
|
Fixed wrong type of arg in: |
| 61 |
|
atm2d/put_ocnvars.F cheapaml/cheapaml.F ocn_compon_interf/ocn_apply_import.F |
| 62 |
|
ctrl/ctrl_map_ini_ecco.F seaice/seaice_do_diags.F seaice/seaice_model.F |
| 63 |
|
To fix: global_sum of var in common block is wrong (multi-threaded) in: |
| 64 |
|
dic/dic_atmos.F ebm/ebm_area_t.F ebm/ebm_zonalmean.F sbo/sbo_calc.F |
| 65 |
|
To fix: aim_v23/aim_do_co2.F (BARRIER & GLOBAL_SUM within bi,bj loops) |
| 66 |
|
o adjustment.cs-32x32x1: test "blank tiles" when compiling with mpi; 2nd |
| 67 |
|
"nlfs" mpi test is going to stop when checking for unconnected tile egdes. |
| 68 |
|
o eesupp/src: new S/R to stop everyone when 1 (or more) Proc. find an error |
| 69 |
|
o pkg/exch2: new S/R EXCH2_CHECK_DEPTHS to check that unconnected tile |
| 70 |
|
edges are closed (zero depth); might happen with "blank tiles"; |
| 71 |
|
for now, disabled if using OBCS. |
| 72 |
|
o tools/gemake2: delete modifs from March 27-29 2009 (with option "-mpi": |
| 73 |
link *_mpi files from "MODS" dir) to return to a strait-forward behavior. |
link *_mpi files from "MODS" dir) to return to a strait-forward behavior. |
| 74 |
o move read_write.F to write_utils.F ; new parameter (maxLengthPrt1D) |
o move read_write.F to write_utils.F ; new parameter (maxLengthPrt1D) |
| 75 |
to reduce length of 1.D array print in STDOUT. |
to reduce length of 1.D array print in STDOUT. |
| 76 |
o adjustment.cs-32x32x1: changed to oceanic test with 1 continent |
o adjustment.cs-32x32x1: changed to oceanic test with 1 continent |
| 77 |
+ use 48 tiles (16x8): this will allow to test "blank tiles". |
+ use 48 tiles (16x8): this will allow to test "blank tiles". |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch