[Mpi3-hybridpm] Changes to ticket #154

[Pavan Balaji]

Hi all,

I've made some changes to the description of ticket #154 (MPI Progress 
Semantics) based on the feedback from the last telecon. There are two 
primary changes:

1. There's a first attempt on describing what "MPI progress semantics" 
means.

2. I've decoupled the issue of progress in different cases (e.g., 
derived datatypes, wild-card tags/sources) from this ticket; the MPI 
implementation will need to figure out whether the application is using 
these features in a different way (e.g., through environment variables 
passed by the user).

Please take a look at the changes, if you get a chance.

Thanks,

  -- Pavan

-- 
Pavan Balaji
http://www.mcs.anl.gov/~balaji