[Mpi3-hybridpm] Hybrid WG meeting at the Forum

[Douglas Miller]

While the implementation I'd be making would probably not distinguish
between MULTIPLE,  SERIAL, and FUNNEL, I still believe that there are
implementations that do benefit from using a lesser threading model and
would still be able to use helper threads. I think it is short-sighted to
rule out those modes for helper threads.


_______________________________________________
Douglas Miller                  BlueGene Messaging Development
IBM Corp., Rochester, MN USA                     Bldg 030-2 A410
dougmill at us.ibm.com               Douglas Miller/Rochester/IBM


                                                                           
             Pavan Balaji                                                  
             <balaji at mcs.anl.g                                             
             ov>                                                        To 
                                       Douglas Miller/Rochester/IBM at IBMUS  
             12/07/2010 08:08                                           cc 
             AM                        mpi3-hybridpm at lists.mpi-forum.org   
                                                                   Subject 
                                       Re: [Mpi3-hybridpm] Hybrid WG       
                                       meeting at the Forum                
                                                                           
                                                                           
                                                                           
                                                                           
                                                                           
                                                                           





On 12/07/2010 08:03 AM, Douglas Miller wrote:
> Well, I was thinking we did not need to restrict MPI_HELPER_* to only
> MPI_THREAD_MULTIPLE, as I think these routines are easily made thread
> safe independently of the rest of MPI. I would like to see these be
> more-widely usable.

Convincing the Forum about that is an uphill battle, AFAICT (not sure
which way I'll vote on that, either). I'd like to not add that in the
initial proposal. If you like, we can create a second proposal chapter
(like what we are doing in the RMA WG) with these "additional changes"
which will be good to have, but not critical.

  -- Pavan

--
Pavan Balaji
http://www.mcs.anl.gov/~balaji
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mpi-forum.org/pipermail/mpiwg-hybridpm/attachments/20101207/b65b2bb4/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: graycol.gif
Type: image/gif
Size: 105 bytes
Desc: not available
URL: <http://lists.mpi-forum.org/pipermail/mpiwg-hybridpm/attachments/20101207/b65b2bb4/attachment-0003.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pic14720.gif
Type: image/gif
Size: 1255 bytes
Desc: not available
URL: <http://lists.mpi-forum.org/pipermail/mpiwg-hybridpm/attachments/20101207/b65b2bb4/attachment-0004.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ecblank.gif
Type: image/gif
Size: 45 bytes
Desc: not available
URL: <http://lists.mpi-forum.org/pipermail/mpiwg-hybridpm/attachments/20101207/b65b2bb4/attachment-0005.gif>