Parent Directory
| 
Revision Log
| 
Revision Graph
| Links to HEAD: | (view) (download) (annotate) |
| Sticky Tag: |
download today code into "MITgcm_today" (instead of just "MITgcm")
to send testreport output, use new email address: jm_c@mitgcm.org (replacing jmc@dev.mitgcm.org)
fix typo
fix previous modif
before making a copy, check that code has been updated
more convenient saving of previous output
Send testreport results to jmc@dev.mitgcm.org (instead of jmc@mitgcm.org)
adjust for changes in Gael's verification_other experiment (global_oce_cs32).
adjust for changes in Gael's verification_other experiment (global_oce_cs32).
fix previous modif
adjusted for modified shelfice_remeshing experiment
from Contrib verification_other: - add experiment "global_oce_cs32" to gfortran+mpi test - add experiment "global_oce_biogeo_bling" to ifort test
exp. "global_oce_biogeo_bling" from Contrib/verification_other added to standard gfortran daily test
remove "shelfice_remeshing" additional experiment from ifort test
add additional experiment (from contrib/verification_other): shelfice_remeshing to non-MPI gfortran & ifort testing
simplify baudelaire testing script: single one, use with argument 'a' or 'b'
add 2 experiments (atm_gray & offline_cheapaml) from Contrib:verification_other to mpi+mth (gfortran) test and ifc test
save previous summary (before removing MITgcm testing dir)
- run 2 test at a time: no significant slow down (since enough cpu) and get the full set of test done earlier - switch the ifort test to "-devel" ; run the OpenAD test everyday.
small improvement for cases it's run again without cleaning
add 1 weekly (Sunday) OpenAD test on baudelaire (using gfortran, non MPI)
specify makedepend for some testreport test (so that other than default tools/xmakedpend are also tested).
fix an unnecessary cleaning
remove un-needed cvs-update
improve cleaning
remove the hack related to GFORTRAN_OPTFILE env. variable
improve report to log file
once a week, run a TLM test (gfortran -devel, no mpi)
a hack to get special option to compile pkg/fizhi with gfortran
- use "-devel" testreport option for standard (gfortran) testing on baudelaire - use same optfile for MPI and non-MPI tests.
- mpi+mth test on baudelaire: uses a max of 3 mpi-procs - pgi on faulks: run testreport only once
minor editing changes
updated after changing testreport (replace TR_NPROC instead of XX in MPI command)
try MPI tests with variable number of Procs (max = 6)
add 3 new tests with MPI+gfortran (fwd, ad, mpi+mth)
reference platform+compiler is now baudelaire+gfortran (only linux_amd64 gfortran test are expected to have 13 digits of matching)
- prepare for reference output switch (from faulks:g77 to baudelaire:gfortran): relax matching of g77 test to 10 ; default matching with gfortran. - add a daily AD test with g77 ; switch optfile (ifort11) for ifort testing.
add testing script for baudelaire ; remove the one for hugo (dead)
This form allows you to request diffs between any two revisions of this file. For each of the two "sides" of the diff, select a symbolic revision name using the selection box, or choose 'Use Text Field' and enter a numeric revision.
| ViewVC Help | |
| Powered by ViewVC 1.1.22 |