I could live with "implementation defined", but I don't think we tend to use that phrasing. When MPI uses the words "undefined", they tend to mean "the implementation may return the results of /dev/random". From: mpiwg-rma [mailto:[email protected]] On Behalf Of Jeff Hammond Sent: Tuesday, August 27, 2013 3:20 PM To: MPI WG Remote Memory Access working group Subject: Re: [mpiwg-rma] FW: [Mpi3-rma] [EXTERNAL] Re: MPI-3 UNIFIED model updates How does what you want not meet the definition of "implementation defined" rather than "defined"? I know this is semantic hair-splitting, but you are not proposing to define the semantics of RMA in the way that one normally thinks of the term "defined". I don't see how your "definition" is any more of a definition than saying that one can either use MPI_RECV or "some combination of function calls and instructions that would allow the user to receive the data and set the message queue in the proper state as if it were RECVed" since what you are saying we're going to "define" is the user can employ MPI_WIN_SYNC or "some combination of function calls and instructions that would allow the user to synchronize the public and private windows in the same manner as WIN_SYNC would do". Jeff On Tue, Aug 27, 2013 at 11:38 AM, Underwood, Keith D <[email protected]<mailto:[email protected]>> wrote: Yes, I want it to be defined. I don't think that version of defined is a problem. The definition just has to happen in terms of what is in memory and say "an implementation might reorder your loads in fun and exciting ways" ;-) This is pretty close to just saying the public and private window are the same. Undefined in the way MPI phrases it allows a unified where it is impossible to use it without the MPI_WIN_SYNC. That isn't ok. From: mpiwg-rma [mailto:[email protected]<mailto:[email protected]>] On Behalf Of Jeff Hammond Sent: Tuesday, August 27, 2013 2:31 PM To: Pavan Balaji Cc: MPI Forum Subject: Re: [mpiwg-rma] FW: [Mpi3-rma] [EXTERNAL] Re: MPI-3 UNIFIED model updates Keith's definition is not a definition in the sense of MPI semantics being "defined". What he is proposing is to be pedantic about the fact that the user may want to rely upon implementation-defined behavior to realize the best performance or to avoid changing existing SHMEM code. The MPI standard isn't defining something if it's not explicitly defined in the standard. Therefore, it must be called "undefined". If you or Keith want to call this "defined" behavior, then I'm going to insist that it be defined in the usual sense, which means the text of the standard will enumerate the behavior of MPI in an architecture-independent way. Jeff On Tue, Aug 27, 2013 at 11:19 AM, Pavan Balaji <[email protected]<mailto:[email protected]>> wrote: On 08/27/2013 12:57 PM, Jeff Hammond wrote: The MPI standard will not define by what mechanism the SHMEM implementation will establish the necessary consistency. This is what I mean by undefined = implementation-defined. I do not think that Keith is proposing to enumerate all processor memory models and the appropriate fences required to achieve the affect of win_sync but at lower cost. Thus, it is the implementation - or perhaps more accurately, the platform on which the implementation resides - that will define these semantics. Here's what I understood from Keith -- He wants a statement in the standard that says the application can do platform-specific memory consistency to synchronize the public and private windows in UNIFIED. I do agree that that's a very flimsy statement and doesn't really mean much. But I'm calling it a definition, while I think you are saying that it's pretty much undefined. I like leaving it totally undefined better than the above "definition". -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji -- Jeff Hammond [email protected]<mailto:[email protected]> _______________________________________________ mpiwg-rma mailing list [email protected]<mailto:[email protected]> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-rma -- Jeff Hammond [email protected]<mailto:[email protected]>