| 4 |
Notes on tags used in MITgcmUV |
Notes on tags used in MITgcmUV |
| 5 |
============================== |
============================== |
| 6 |
|
|
| 7 |
|
o verification/tutorial_tracer_adjsens: |
| 8 |
|
- OpenAD test: switch to vector-invariant momentum since fluxform with |
| 9 |
|
rStar causes problems with OpenAD; update output. |
| 10 |
|
o tools: |
| 11 |
|
- genmake2: define a common prefix for all temp-file names ; move them |
| 12 |
|
to local /tmp disk (but would be easy to move them back); |
| 13 |
|
- xmakedepend: apply the same switch for temp-files. |
| 14 |
|
o pkg/offline: |
| 15 |
|
- changes for use with rStar (NLFS): assume that we continue to provide |
| 16 |
|
UVELMASS/hFacW_ini & VVELMASS/hFacS_ini (hFacWi,S_ini as written in grid |
| 17 |
|
output files) and apply r* scaling at each time-step (function of the time |
| 18 |
|
varying Eta field); use normal pickup files (needed for restart with NLFS). |
| 19 |
|
o adjoint with OpenAD: |
| 20 |
|
- move OpenAD-specific files from verification/OpenAD/code_oad_all to a new |
| 21 |
|
package openad. This package has to be compiled for all runs with OpenAD |
| 22 |
|
instead of specifying code_oad_all as a modification directory. |
| 23 |
|
- link the fortran templates from tools/OAD_support to the build directory |
| 24 |
|
for preprocessing to avoid conflicts between different setups compiled |
| 25 |
|
simultaneously and simplify cleanup. |
| 26 |
|
o pkg/exf: |
| 27 |
|
- in exf_getforcing.F: comment out most of the EXCH calls (fields used by |
| 28 |
|
main model or other pkgs are exchanged in EXF_MAPFIELDS). |
| 29 |
|
o model/src: |
| 30 |
|
- do_stagger_fields_exchanges.F: skip EXCH calls when using pkg/offline |
| 31 |
|
(similar to do_fields_blocking_exchanges.F). |
| 32 |
|
o tools/genmake2: |
| 33 |
|
- remove TAF output src code just before calling staf (when not doing a fresh |
| 34 |
|
built, will stop if things go wrong with staf). |
| 35 |
|
- fix target "adobj" (left from TAF version-2 transition) |
| 36 |
|
o pkg/diagnostics: |
| 37 |
|
- add a parameter to track the status of pkg/diagnostics activation; |
| 38 |
|
this replace/extend the use of logical param "settingDiags" (to check |
| 39 |
|
when adding diag to the list) to also check any DIAGNOSTICS_[]FILL* call. |
| 40 |
|
New S/R to print error message and stop when pkgStatus is not right. |
| 41 |
|
- to try: add a warning when "data.diagnostics" exist but useDiagnostics=F. |
| 42 |
|
|
| 43 |
|
checkpoint64m (2013/08/13) |
| 44 |
|
o model/inc: |
| 45 |
|
- move FCORI_MACROS.h from PARAMS(_MACROS).h to GRID_MACROS.h (left from |
| 46 |
|
8 yrs ago modif when Coriolis arrays were moved from PARAMS.h to GRID.h) |
| 47 |
|
o pkg/kpp & aim_v23: |
| 48 |
|
- only call DIAGNOSTICS_FILL when useDiagnostics=TRUE. |
| 49 |
|
o pkg/seaice: |
| 50 |
|
- restore (last time on 2011/12/24) ability to compile pkg/seaice with |
| 51 |
|
undefined pkg/exf ALLOW_ATM_TEMP or ALLOW_RUNOFF CPP-options. |
| 52 |
|
- only call DIAGNOSTICS_FILL when useDiagnostics=TRUE. |
| 53 |
|
o eesupp/src: |
| 54 |
|
- fix check for overlap-size argument (myOLn.GT.MAX_OLY_EXCH) |
| 55 |
|
in exch1_rx[_ad].template (thanks to Francois Lefeuvre report). |
| 56 |
|
o switch to TAF version-2: |
| 57 |
|
- add flow directives to "freeze" naming convention for ad-routines that |
| 58 |
|
are used in I/O routines of adjoint state and associated common blocks |
| 59 |
|
that are used in adcommon.h (backward compatible with version-1) |
| 60 |
|
- update all ad-optfiles (in tools/adjoint_options + few local to some |
| 61 |
|
verification exp) to switch to latest TAF version (currently 2.5.5); |
| 62 |
|
merge "*_mpi" adoptfiles into the corresponding non-mpi (but commented |
| 63 |
|
out except for diva). |
| 64 |
|
- the major change with TAF version-2 is that both the differentiated |
| 65 |
|
routines and the forward routines (new) are returned as part of the TAF |
| 66 |
|
output source code. Update genmake2 accordingly. |
| 67 |
|
- new output for AD experiment tutorial_global_oce_biogeo (better AD gradient |
| 68 |
|
with version-2 when compared to TLM or FWD gradient); other experiments |
| 69 |
|
output show minor changes (machine truncation level); update the few AD & |
| 70 |
|
TLM output with 12 matching digits only (some prior to version-2 switch). |
| 71 |
|
|
| 72 |
|
checkpoint64l (2013/08/05) |
| 73 |
|
o model/src: |
| 74 |
|
- in dynamics.F, skip the call to CALC_VISCOSITY if momViscosity=F |
| 75 |
|
- main_do_loop.F: no longer includes additional (AUTODIFF) header files |
| 76 |
|
when using TAF but only if using OpenAD. |
| 77 |
|
o pkg/mom_common+fluxform+vecinv: |
| 78 |
|
- always set horiz. viscosity arrays to background value before calling |
| 79 |
|
MOM_CALC_VISC (in MOM_FLUXFORM & MOM_VECINV) and call S/R MOM_CALC_VISC |
| 80 |
|
only when using variable horiz. viscosity (useVariableVisc=T); |
| 81 |
|
- simplify mom_vecinv.F (only 1 block for momViscosity). |
| 82 |
|
o pkg/mom_common: variable horiz. viscosity in NH code |
| 83 |
|
- move horiz-viscosity 3-D arrays viscAh_W & viscA4_W (used in calc_gw.F) |
| 84 |
|
from model/inc/NH_VARS.h to pkg/mom_common/MOM_VISC.h |
| 85 |
|
- store in common block (in MOM_VISC.h) logical parameters: useHarmonicVisc, |
| 86 |
|
useBiharmonicVisc & useVariableVisc. |
| 87 |
|
- fix old bug (already in 1rst implementation of variable horizontal viscosity |
| 88 |
|
in calc_gw.F, version 1.27, 2006/06/20): |
| 89 |
|
horizontal bi-harmonic viscosity in gW was missing when wiscA4W=0 |
| 90 |
|
even if variable bi-harmonic viscosity (e.g., viscA4Grid) was non-zero. |
| 91 |
|
- change background horiz viscosity in vertical momentum (gW) to always be |
| 92 |
|
viscAhW,viscA4W (instead of switching to viscAhD, viscA4D when using |
| 93 |
|
any type of variable viscosity, e.g., viscAh/4Grid) |
| 94 |
|
- change default value of viscAhW,viscA4W to viscAhD,viscA4D (instead of |
| 95 |
|
viscAh,viscA4) which default to viscAh,viscA4 : this avoid missing |
| 96 |
|
horiz. viscosity in W eq when just setting viscAhD/Z,viscA4D/Z (since |
| 97 |
|
default viscAh,viscA4 is zero). |
| 98 |
|
o model/src: |
| 99 |
|
- extend length of drC & recip_drC from Nr to Nr+1. |
| 100 |
|
o Divided Adjoint (DIVA) with MPI: |
| 101 |
|
- genmake2: do not create local ./mpi_headers dir when no file to copy into. |
| 102 |
|
- remove option ALLOW_DIVIDED_ADJOINT_MPI (replaced by ALLOW_DIVIDED_ADJOINT |
| 103 |
|
and ALLOW_USE_MPI) from model/src code and from ECCO_CPPOPTIONS.h. |
| 104 |
|
|
| 105 |
|
checkpoint64k (2013/07/23) |
| 106 |
|
o tools/OAD_support: |
| 107 |
|
- avoid the need for file "oadtempflile" in insertTemplateDir.bash |
| 108 |
|
and remove it from genmake2. |
| 109 |
|
o tools/build_options: |
| 110 |
|
- remove MPI_HEADER_FILES_INC setting (no longer used) and comment out |
| 111 |
|
MPI_HEADER_FILES list (let genmake2 find the missing headers) |
| 112 |
|
o verification/halfpipe_streamice: |
| 113 |
|
- change to driving stress formulation; update both Fwd & AD output. |
| 114 |
|
o tools/genmake2 and Divided Adjoint with MPI: |
| 115 |
|
- remove default setting of MPIINCLUDEDIR from $MPIHOME (always been broken); |
| 116 |
|
- only create local ./mpi_headers dir if using MPI and MPIINCLUDEDIR is set; |
| 117 |
|
- add new variable (LOCAL_MPI_HEADERS, replace MPI_HEADER_FILES_INC) in |
| 118 |
|
Makefile, which is set by genmake2, and check which mpi-headers are needed. |
| 119 |
|
Note: need to update adjoint_options/adjoint_diva_mpi accordingly. |
| 120 |
|
- update target "CLEAN" (to also remove ./mpi_headers/ dir). |
| 121 |
|
o testreport: |
| 122 |
|
- changes for Divided Adjoint Run: when file "run_ADM_DIVA" is present, |
| 123 |
|
perform additional runs (number of additional runs is "add_DIVA_runs" taken |
| 124 |
|
from file run_ADM_DIVA ; should match "nchklev_3" in file "code_ad/tamc.h"). |
| 125 |
|
o verification/lab_sea (AD test): |
| 126 |
|
- change to test Divided Adjoint Run in this set-up (both standard AD test |
| 127 |
|
and 2 secondary AD test: noseaice + noseaicedyn); drop the evp AD test. |
| 128 |
|
o pkg atm_ & ocn_compon_interf: |
| 129 |
|
- remove cpl_earlyExpImpCall=F option (was mainly a hack added for the old |
| 130 |
|
staggerTimeStep implementation) and disable (retired) corresponding |
| 131 |
|
parameter. |
| 132 |
o pkg/offline: |
o pkg/offline: |
| 133 |
- comment out unused field-array: ICEM, ConvectCount, hflx[01], icem[01]; |
- comment out unused field-array: ICEM, ConvectCount, hflx[01], icem[01]; |
| 134 |
- add S/R OFFLINE_CHECK to print offline parameter summary |
- add S/R OFFLINE_CHECK to print offline parameter summary |
| 135 |
and stop if loading a file that is not used; |
and stop if loading a file that is not used; |
| 136 |
- enable to calculate GMRedi tensor, KPP mixing and convective index |
- enable to calculate GMRedi tensor, KPP mixing and convective index |
| 137 |
when the corresponding file(s) is/are not specified (but pkg is used). |
when the corresponding file(s) is/are not specified (but pkg is used). |
| 138 |
|
- comment out resetting of monitorFreq (to endTime) since it still useful |
| 139 |
|
for monitoring CFL numbers and checking offline fields. |
| 140 |
o OpenAD: |
o OpenAD: |
| 141 |
- do not define OAD_DEBUG (in tools/adjoint_options/adjoint_oad) to get |
- do not define OAD_DEBUG (in tools/adjoint_options/adjoint_oad) to get |
| 142 |
smaller standard output) and generate new (smaller) output (on baudelaire |
smaller standard output) and generate new (smaller) output (on baudelaire |
Parent Directory
| 
Revision Log
| 
Revision Graph
| 
Patch