Yes that seems to fix my problem. Sorry for the false alarm. I need to
remember to look a litter farther down in the MPI Reference manual when
looking at the function declarations.
On Aug 30, 2005, at 7:30 AM, Jeff Squyres wrote:
> I think you forgot the ierr argument in your call to MPI_WAITALL. This
> is a common case for a segv in a fortran application (missing function
> arguments, leading to bizarre side-effects such as yours [works "fine"
> unless you add the write statement]).
>
>
> Can you try adding that ", ierr" and see if that makes the problem go
> away?
>
>
> On Aug 25, 2005, at 3:47 PM, Douglas Vechinski wrote:
>
>
> > I posted a message a week ago about a problem I am having but didn't
> > hear responses. So I'm posting it again with an example that is giving
> > me problems. Below is a reiteration of the problem. I've attached two
> > small Fortran codes which demonstrate the problem (at least for me).
> > The first mst.f is the master which spawns a couple of slave processes
> > (slv.f). Make sure the slave executable is called 'slv' since this is
> > the name that is used to spawn it off with. The master is set up to
> > spawn two slaves off. Execute the master with only one process
> >
>
>
|
