You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I found this from working with izumi, which uses the mvapich2 MPI library for all compilers: intel, gnu, and nag.
There's a call in mpi_process.f90 subroutine pass_global_data to MPI_ADDREDUCE and it sends and receive the same
integer variable: maxtdh. It just needs the send and receive buffer to be different.
The cesm.log gives this kind of error:
[0] (shr_orb_params) ------ Computed Orbital Parameters ------
[0] (shr_orb_params) Eccentricity = 1.670366E-02
[0] (shr_orb_params) Obliquity (deg) = 2.343977E+01
[0] (shr_orb_params) Obliquity (rad) = 4.091011E-01
[0] (shr_orb_params) Long of perh(deg) = 1.028955E+02
[0] (shr_orb_params) Long of perh(rad) = 4.937458E+00
[0] (shr_orb_params) Long at v.e.(rad) = -3.247250E-02
[0] (shr_orb_params) -----------------------------------------
[[email protected]] HYDT_bscd_pbs_wait_for_completion (tools/bootstrap/external/pbs_wait.c:67): tm_poll(obit_event) failed with TM error 17002
[[email protected]] HYDT_bsci_wait_for_completion (tools/bootstrap/src/bsci_wait.c:23): launcher returned error waiting for completion
The text was updated successfully, but these errors were encountered:
I found this from working with izumi, which uses the mvapich2 MPI library for all compilers: intel, gnu, and nag.
There's a call in mpi_process.f90 subroutine pass_global_data to MPI_ADDREDUCE and it sends and receive the same
integer variable: maxtdh. It just needs the send and receive buffer to be different.
The cesm.log gives this kind of error:
The text was updated successfully, but these errors were encountered: