Opened 14 years ago
Closed 14 years ago
#514 closed bug (Fixed)
mpi exchange in ana_wwave — at Version 1
Reported by: | abever | Owned by: | arango |
---|---|---|---|
Priority: | major | Milestone: | Release ROMS/TOMS 3.5 |
Component: | Nonlinear | Version: | 3.5 |
Keywords: | Cc: |
Description (last modified by )
We were having problems with test cases using the SSW_BBL bottom boundary layer formulation when running in parallel. This setup created irregularities along the tile boundaries.
It looks like mp_exchange was not always called from ana_wwave.h. It looks like mp_exchange was inside the EW_PERIODIC or NS_PERIODIC C-preprocessing conditional. Thanks to John for figuring this out. A file that works is attached.
Change History (2)
by , 14 years ago
Attachment: | ana_wwave.h added |
---|
comment:1 by , 14 years ago
Description: | modified (diff) |
---|---|
Resolution: | → Fixed |
Status: | new → closed |
Yes, good catch!!! This may explain a parallel bug that is in my to do list when the bottom boundary layer formulation is activated with analytical options for the wind-induced wave forcing. Thank you for bringing this to our attention.
Working ana_wwave.h