[Mpi3-rma] Conflicting PUT/GET accesses clarification
James Dinan
dinan at mcs.anl.gov
Wed Apr 27 17:12:24 CDT 2011
Logically, I would think of this as two concurrent (conflicting) updates
to the same location. The MPI implementation could detect and avoid a
problem. The same goes for two conflicting puts; the MPI implementation
can also detect and avoid the problem but we define it to be erroneous.
How is this defined for send/recv? I think I looked it up once and
found that it was an error in that case.
~Jim.
On 04/27/2011 05:07 PM, Pavan Balaji wrote:
>
> It's not two concurrent accesses -- it's a single (overlapping) access.
>
> -- Pavan
>
> On 04/27/2011 04:58 PM, James Dinan wrote:
>> On 04/27/2011 03:51 PM, Pavan Balaji wrote:
>>>
>>> Page 339 lines 21-24 of MPI-2.2 states the following:
>>>
>>> "It is erroneous to have concurrent confl icting accesses to the same
>>> memory location in a window; if a location is updated by a put or
>>> accumulate operation, then this location cannot be accessed by a load or
>>> another RMA operation until the updating operation has completed at the
>>> target."
>>>
>>> What does this mean when I do PUT with a datatype with overlapping
>>> buffers? Is that valid or erroneous?
>>
>> My interpretation is that this is also erroneous.
>>
>> ~Jim.
>> _______________________________________________
>> mpi3-rma mailing list
>> mpi3-rma at lists.mpi-forum.org
>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
>
More information about the mpiwg-rma
mailing list