Hi,
I checked the user document and MAN pages yet.
But,,, regrettable,,, I can't understand the usage of cr self module.
If a code is ...
-------------------------------------------------
int cr_restart(void)
{
std::cout << "test restart" << std::endl;
return 0;
}
main(int argc, char* argv[])
{
/* Start up MPI */
int provided = MPI::Init_thread(argc, argv, MPI::THREAD_SERIALIZED);
//some MPI procedure...
while(false){
cr_restart();
}
//some MPI procedure...
/* Shut down MPI */
MPI::Finalize();
} /* main */
-------------------------------------------------
and
$ mpic++ -O0 test.cpp
$ lamboot -ssi rpi crtcp
$ lamrestart -ssi cr self -ssi cr_self_restart_args "-ssi cr_self_prefix cr -np 2 ./a.out"
.
Should an output prompt show "test restart" ?
In addition... man page for lamrestart and lamrestart prompt contain typo.
cr_restart_args -> cr_self_restart_args
Thanks.
>>>>> On Fri, 17 Mar 2006 13:22:58 -0500,
>>>>> [afriedle_at_[hidden]] wrote:
> I've found that section 9.5 of the LAM User's Manual discusses use of
> checkpoint/restart:
> http://www.lam-mpi.org/download/files/7.1.2-user.pdf
> Also, some LAM man pages might be of use to you:
> lamssi_cr(7)
> lamcheckpoint(1)
> lamrestart(1)
> These are available with LAM itself and can be found via your favorite
> search engine as well.
> Although the blcr module is likely what you want, the self module looks
> fairly easy to use (I have no personal experience with it myself), in
> that the application defines three functions that are called upon
> checkpoint, restart, and continue. What these functions need to do will
> be specific to your application - I can't help you there.
> Andrew
> _______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
|