*If TEST_NAME specifies WARM_START=.T., error and exit
*Restart file names are automatically determined based on SYEAR, SMONTH, SDAY, SHOUR, FHMAX
*RESTART_INTERVAL automatically determined as one half of FHMAX
*Restart comparison files are automatically determined based on OUTPUT_GRID and OUTPUT_FILE
*MPI testing sets WRITE_GROUP=2 and WRTTASK_PER_GROUP=12
*Name of the lock directory changed from lock to lock_ut to avoid conflict with RT

…conf)

*Add more test-names to utest.bld
*Further modifications to differentiate output file names from those of RT
*RT passed

* Time taken for full unit tests is 40% of of that without using ecflow
* Numerous changes in utest for ecflow as well as better handling of baseline runs
* Modify run_test.sh to handle any failure before run_test
* Modify rt.sh to use ECFLOW_SUITE instead of regtest
* Modify rt_utils.sh to use ECFLOW_SUITE instead of regtest
* In rt_utils.sh, modify ecflow_create_run_task: make the call to run_test a foreground job so that 'set -e' error handling works properly by getting the correct exit status (another way would be to use 'job id' and 'wait job id' following the call to run_test &)
* Modify some of fv3_conf/*_run.IN files to handle errors when restart file copies fail

…nd testing

…retched_nest and fv3_(ccpp_)regional_c768

…nml.IN; this was done in parm/ccpp_input_nest02.nml.IN, but forgotten in parm/input_nest02.nml.IN

…r-model into hafs_sas_for_develop

…/ufs-weather-model into hafs_sas_for_develop

…ufs-weather-model into hafs_sas_for_develop

…t.intel