Custom Query (986 matches)
Results (25 - 27 of 986)
| Ticket | Owner | Reporter | Resolution | Summary | |
|---|---|---|---|---|---|
| #23 | Fixed | Miscellaneous fixes to Functionals files | |||
| Description |
ana_diag.h: Remove declarations for u & v when SOLVE3D is undefined.
Files attached. |
||||
| #25 | Fixed | Declaration problem in cgradient.h | |||
| Description |
In file cgradient.h, subroutine read_state.F, the explicit-shape version of the declaration for argument s_t is real(r8), intent(inout) :: s_t(LBi:UBi,LBj:UBj,N(ng),NT(ng)) This has only 4 dimensions, whereas the corresponding assumed-shape declaration has 5: real(r8), intent(inout) :: s_t(LBi:,LBj:,:,:,:) I think the correct explicit-shape declaration is real(r8), intent(inout) :: s_t(LBi:UBi,LBj:UBj,N(ng),2,NT(ng)) Corrected file attached. |
||||
| #26 | Fixed | ROMS fails to initialize in parallel with point sources/sink | |||
| Description |
I ran the LATTE model for several cases with ROMS 3.0 changeset 56. Compiled on moby using pgf90. If I set tiles to 2x2 (or 1x4) the program freezes on initialization(at runtime). This does not happen with a 1x1 tile or if I compile in serial. Also, the upwelling case compiles and initializes fine at runtime. Once UV_SOURCE and TS_SOURCE are undefined there is no apparent problem running in parallel. |
||||
