Yeah I didn't really know what to set as the default behavior for that function. Does someone have experience using it that could suggest a reasonable value? Thanks, Josh On May 2, 2011, at 12:39 PM, Howard Pritchard wrote:
Hi Folks,
Did we ever decide how mpi_file_set_view would work after the application has MPI_File_validate_sync on the file handle?
I could see some apps wanting 'holes' for missing processes, but then other times it might not want holes.
Could this behavior be handled by some kind of pre-defined info key and then the app would pass that in to MPI_file_set_view after doing the mpi_file_validate_sync? An issue I see with this approach is there's no way to get back what the info value was using MPI_File_get_view.
Howard
-- Howard Pritchard Software Engineering Cray, Inc. _______________________________________________ mpi3-ft mailing list [email protected] http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft