actually I am wondering whether my previous statement was correct. If
you do not intercept the fortran MPI call, than it still goes to the C
MPI call, which you can intercept. Only if you intercept the fortran MPI
call we do not call the C MPI but the C PMPI call, correct? So in
theory, it could still work...
Jeff Squyres wrote:
> On Dec 5, 2008, at 12:22 PM, Edgar Gabriel wrote:
>
>> I hope you are aware, that *many* tools and application actually
>> profile the fortran MPI layer by intercepting the C function calls.
>> This allows them to not have to deal with f2c translation of MPI
>> objects and not worry about the name mangling issue. Would there be a
>> way to have both options e.g. as a configure flag? The current commit
>> basically breaks all of these applications...
>
> I was unaware of this, actually.
>
> So it'd be pretty easy to have a configure switch for this (it would be
> a bunch more work for a run-time switch; I don't know if it's really
> worth it?). Should we default to the current behavior, and have the
> configure switch enable call stacks like this:
>
> MPI_Comm_rank_f
> PMPI_Comm_rank_f
> PMPI_Comm_rank
>
> ?
>
--
Edgar Gabriel
Assistant Professor
Parallel Software Technologies Lab http://pstl.cs.uh.edu
Department of Computer Science University of Houston
Philip G. Hoffman Hall, Room 524 Houston, TX-77204, USA
Tel: +1 (713) 743-3857 Fax: +1 (713) 743-3335
|
![[Flag]](../../../../../community/mirrors/flag-us.gif){kind=small}