I'm trying to run IMB-EXT (part of Intel MPI Benchmark 3.0) with OpenMPI-1.1.2, on FreeBSD 6.2-PRERELEASE on x86.
[trasz@traszkan:~]$ mpirun -np 2 IMB-EXT #--------------------------------------------------- # Intel (R) MPI Benchmark Suite V3.0, MPI-2 part #--------------------------------------------------- # Date : Sun Nov 19 21:33:46 2006 # Machine : i386 # System : FreeBSD # Release : 6.2-PRERELEASE # Version : FreeBSD 6.2-PRERELEASE #1: Mon Oct 30 07:41:21 CET 2006 trasz@traszkan.ds7:/usr/obj/usr/src/sys/TRASHCAN # MPI Version : 2.0 # MPI Thread Environment: MPI_THREAD_SINGLE # # Minimum message length in bytes: 0 # Maximum message length in bytes: 4194304 # # MPI_Datatype : MPI_BYTE # MPI_Datatype for reductions : MPI_FLOAT # MPI_Op : MPI_SUM # # # List of Benchmarks to run: # Window # Unidir_Get # Unidir_Put # Bidir_Get # Bidir_Put # Accumulate [traszkan.ds7:10583] *** An error occurred in MPI_Win_free [traszkan.ds7:10583] *** on win [traszkan.ds7:10583] *** MPI_ERR_RMA_SYNC: error while executing rma sync [traszkan.ds7:10583] *** MPI_ERRORS_ARE_FATAL (goodbye) [traszkan.ds7:10584] *** An error occurred in MPI_Win_free [traszkan.ds7:10584] *** on win [traszkan.ds7:10584] *** MPI_ERR_RMA_SYNC: error while executing rma sync [traszkan.ds7:10584] *** MPI_ERRORS_ARE_FATAL (goodbye) 1 additional process aborted (not shown) This is single CPU machine, not part of any cluster. Running with '-np 1' results in the same problem, only faster. The same test with MPICH2 works without problems. Any clues?
