On Feb 1, 2006, at 11:55 PM, wchao_at_[hidden] wrote:
> Seems the process of lam can not invoke multiple nsend()/nrecv() call
> within multiple threads, or the call may be blocked on
> read(socket-to-the-lam-kernel-process) forever. That's due to the
> single
> socket between the process and the lam kernel process, right?
That is one reason yes -- but a better characterization is that the
LAM layer really was not designed to be thread-safe or thread-
concurrent. So yes, this behavior is expected. :-(
> Anyway, if I modified my new process to run in sequent flow,
> instead of
> multithread, that issue would disappear.
Excellent.
--
{+} Jeff Squyres
{+} The Open MPI Project
{+} http://www.open-mpi.org/
|
