Open MPI logo

Open MPI User's Mailing List Archives

  |   Home   |   Support   |   FAQ   |   all Open MPI User's mailing list

From: Jeff Squyres (jsquyres_at_[hidden])
Date: 2006-03-03 18:31:38


On Mar 1, 2006, at 11:08 AM, Benoit Semelin wrote:

> call MPI_BCAST(boundary_cond,
> 8,MPI_CHARACTER,master,MPI_COMM_WORLD,mpi_err)
> 1
> Error: Generic subroutine 'mpi_bcast' at (1) is not an intrinsic
> subroutine

It looks like we goofed; we neglected to include F90 routines for the
CHARACTER type in the buffer. :-(

> If I use include 'mpif.h', it compiles just fine.

This is because that will use the [exactly equivalent] F77 bindings
-- you just lose the type safety. Until we can fix the bindings, if
you need to use CHARACTER buffer types, this is probably the best
workaround.

> Second topic:
> I am using 3 processors
> I am calling a series of MPI_SCATTER which work when I send
> messages of
> 5 ko to the other processors, fails at the second scatter if I sent
> messages of ~10 ko, and fails at the first scatter for bigger
> messages.
> The message is:

What is "ko" -- did you mean "kb"?

> 2 processes killed (possibly by Open MPI)

> Could this be a problem of maximum allowed message size? Or of
> buffering
> space?

No, Open MPI should allow scattering of arbitrary sized messages.
Can you verify that your arguments to MPI_SCATTER are correct, such
as buffer length, the receive sizes on the clients, etc.?

Are any corefiles generated? Do you know which processes die?

-- 
{+} Jeff Squyres
{+} The Open MPI Project
{+} http://www.open-mpi.org/