Charm++ parallel programming system

[Xoan Carlos Pardo Martinez <xoanpardo AT gmail.com>]

Hi Chee,

Thank you very much for your answer.

I´m still having some doubts on the right way to make my experiments. At this point I´ve compiled CHARM to use

default TCP communications over InfiniBand (so no underlying MPI communications should be used by CHARM).

I´m interested in measuring exclusive/inclusive times of both MPI and non-MPI routines for different programs

(some of them taken from the NPB benchmarks).

What I fail to see is if I really need to use the Charm++-TAU measurement interface. If I compile TAU to use CHARM

threads and AMPI as underlying MPI implementation, build my programs with AMPI and run them by using the TAU 

interposition library, should I get the times as seen by the AMPI application, isn´t it or am I missing anything?

I take note of your offering for interactive assitance, but before contacting you I'd like to make some tests to have

it clear to myself in what I really need assistance  ;-D 

Best regards

----

Xoan C. Pardo

Computer Architecture Group

University of A Coruña - Spain

Em 13/04/2012, às 19:18, Chee Wai Lee escreveu:

Charm++-TAU measurement interface

> Hi Xoan,
>
> 	I do not believe anyone has attempted to use TAU profiling (via the Charm++-TAU measurement interface) on AMPI so far. I should be able to help you navigate through some of the issues and perhaps adapt the framework to enable what you desire if some functionality is missing.
>
> 	The way TAU interfaces with the Charm++ runtime to measure the performance of Charm++ events (on top of which AMPI is built) is through callback hooks provided by the runtime system. You will need to follow steps similar to the following:
>
> 	http://www.nic.uoregon.edu/tau-wiki/Guide:NAMDTAU
>
> 	The only requirements on the configuration and building of TAU is that it should match the Charm++ build as closely as possible (the architecture, the compilers and the use of any underlying MPI layers). In your case, the use of "-no-trace-mpi" is critical. As a default, TAU wraps MPI calls. In the case of attempting to measure AMPI performance, we do not want to do that (it is actually unclear which MPI layer gets wrapped in this case) but instead interpret the calls according to what the Charm++ runtime tells TAU.
>
> 	You may contact me via skype if you need any interactive assistance on the matter. What you are doing is new as far as I know and might require some work to get going correctly. Thanks!
>
> Best Regards,
> Chee Wai Lee
> University of Oregon