charm AT lists.siebelschool.illinois.edu

Subject: Charm++ parallel programming system

List archive

Re: [charm] Charm++ and MPI

From: Phil Miller <mille121 AT illinois.edu>
To: François Tessier <francois.tessier AT inria.fr>
Cc: Emmanuel Jeannot <emmanuel.jeannot AT inria.fr>, "charm AT cs.uiuc.edu" <charm AT cs.uiuc.edu>, Guillaume Mercier <mercier AT labri.fr>
Subject: Re: [charm] Charm++ and MPI
Date: Fri, 31 Jan 2014 14:54:38 -0600
List-archive: <http://lists.cs.uiuc.edu/pipermail/charm/>
List-id: CHARM parallel programming system <charm.cs.uiuc.edu>

On Fri, Jan 31, 2014 at 2:37 PM, Nikhil Jain <nikhil.jain AT acm.org> wrote:

- If I create a new communicator with differents MPI ranks, can I
apply it during the execution? How does it work exactly?

Do you intend to use MPI calls inside the Charm code? It will work as
usual MPI calls, but you have to be very careful about deadlocks -
blocking MPI calls may prevent execution of other chares on a process
that may lead of deadlocks.

Our documentation on Charm++-MPI interoperability might be useful reading as well:
http://charm.cs.illinois.edu/manuals/html/charm++/25.html

[charm] Charm++ and MPI, François Tessier, 01/30/2014
- Re: [charm] Charm++ and MPI, Nikhil Jain, 01/31/2014
  - Re: [charm] Charm++ and MPI, Phil Miller, 01/31/2014