[Mpi3-rma] MPI-3.1 consideration slides

Pavan Balaji balaji at mcs.anl.gov
Tue Dec 4 13:48:46 CST 2012


Ok.  Very unclear, though.  Some better language won't hurt.

 -- Pavan

On 12/04/2012 01:41 PM US Central Time, Rajeev Thakur wrote:
> Yes, that very sentence says that :-). Read it like this:
> 
> "A put or accumulate must not access a target window once a load/store update -- or a put or accumulate update to another overlapping target window -- has started on a location in the target window, until the update becomes visible in the public copy of the window.    ..... the restriction applies to operations even if they access disjoint locations in the window."
> 
> 
> On Dec 4, 2012, at 1:35 PM, Pavan Balaji wrote:
> 
>>
>> On 12/04/2012 01:33 PM US Central Time, Rajeev Thakur wrote:
>>> I am reading it as "load/store update to the same window" and  "put
>>> or accumulate update to a different overlapping window". That is
>>> because the next sentence says "the restriction applies to operations
>>> even if they access disjoint locations in the window." That
>>> restriction cannot apply to concurrent puts and accumulates to the
>>> same window.
>>
>> Maybe this is an interpretation thing, but I'm looking for a statement
>> that says "concurrent load/store accesses and PUT/GET operations to the
>> same window is not valid".  Am I missing it somewhere?
>>
>> -- Pavan
>>
>> -- 
>> Pavan Balaji
>> http://www.mcs.anl.gov/~balaji
> 

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



More information about the mpiwg-rma mailing list