Bug fix for FMS2 issue #761, broadcast from root pe #1462
Conversation
- This addresses the FMS issue $761 NOAA-GFDL/FMS#761 - There is a mpp_broadcast in the FMS2 subroutine get_unlimited_dimension_name() and this subroutine has to be called by all pes, so it cannot be inside a if(is_root_pe()) block
Codecov Report
@@ Coverage Diff @@
## dev/gfdl #1462 +/- ##
=========================================
Coverage 29.13% 29.13%
=========================================
Files 235 235
Lines 71039 71038 -1
=========================================
Hits 20694 20694
+ Misses 50345 50344 -1
Continue to review full report at Codecov.
|
There was a problem hiding this comment.
Although I agree that this PR fixes a problem that was introduced by FMS2, I think that we do not want any processors apart from the root PE issuing any warning messages, so that we are not deluged with a flood of redundant messages. Please add a second expression to the logical test that is now on line 1102, so that it reads if ((t1 /= t2) .and. is_root_PE()) then, to avoid excessive numbers of warnings.
- This is restricting the warning message in MOM_restart.F90 to root pe per Bob.
|
Gaea regression: https://gitlab.gfdl.noaa.gov/ogrp/MOM6/-/pipelines/13336 ✔️ |
This addresses the FMS issue $761
FMS2+MOM6 model crashes during restart with mpich errors NOAA-GFDL/FMS#761
There is a mpp_broadcast in the FMS2 subroutine
get_unlimited_dimension_name() and this subroutine has to be called by
all pes, so it cannot be inside a if(is_root_pe()) block