Agreed. On 6/3/15, 8:18 PM, "Balaji, Pavan" <[email protected]> wrote:

Yes, I spoke loosely, but what I said doesn't change with that distinction. "Matched" is a precursor to "received".

"If the user never posts a receive, the send cancel should always succeed."

However, if you interpret "received at the destination" to include unexpected queues, then one should not be able to cancel the request.

My interpretation is the former one. But for either interpretation, the standard must be clarified.

-- Pavan

On 6/3/15, 8:13 PM, "Sur, Sayantan" <[email protected]> wrote:

...

The standard distinguishes between “received” and “matched” very carefully. For example, MPI_Ssend() can return when the message is matched but not necessarily fully received.

"A send that uses the synchronous mode can be started whether or not a matching receive was posted. However, the send will complete successfully only if a matching receive is posted, and the receive operation has started to receive the message sent by the synchronous send."

Therefore, if the standard says “received”, I do not want to interpret that as “matched”.

On 6/3/15, 8:03 PM, "Balaji, Pavan" <[email protected]> wrote:

...

If you interpret received at the destination as "matched", then what I mentioned below holds -- if the user never posts a receive, the send cancel should always succeed.

-- Pavan

On 6/3/15, 7:59 PM, "Sur, Sayantan" <[email protected]> wrote:

...

+MPI P2P WG.

We should probably discuss it in the P2P WG as well.

I read the snippet of text and thought the “send” mentioned should correspond to the MPI send operation (the one the user issued), and not any internal buffering. “send completes normally” should mean the same thing as sends completing as if cancel wasn’t called. Similarly, “received at the destination” should mean the normal way messages are received at the destination i.e. by posting an MPI receive. Of course, these are my interpretations :)

On 6/3/15, 7:19 PM, "Balaji, Pavan" <[email protected]> wrote:

...

[Cc'ed the Forum]

This wording has been debated many times at the Forum. Specifically, the "completes normally" and "received at the destination" pieces are ambiguous. For instance, if the data is copied into a bounce buffer, the send would complete normally, but that does not mean that there is a receive posted for that data. In this case, it would be the user's responsibility to post a corresponding receive. If the MPI implementation does not have a bounce buffer (e.g., for large messages), "received at the destination" could mean into an unexpected buffer, not necessarily a user buffer.

The consensus from MPI-1 veterans (Marc, Bill, etc.) during the last discussion was that if the application does not post a receive, then it is guaranteed that the send should be cancelable. At least that was the intent. The wording, of course, is too vague for such a definition.

You could try to weaken that interpretation and say that if the MPI implementation is not able to cancel a send operation, then the user is responsible for posting a matching receive to receive that data. Irrespective of which interpretation we go with, I think it is important to clarify it in the standard. Perhaps an MPI-3.1 errata or MPI-3.2 change instead of waiting for MPI-4? Do you want to create a ticket?

-- Pavan

On 6/3/15, 7:05 PM, "Sur, Sayantan" <[email protected]> wrote:

...

What lines in the standard support the interpretation that “a message that will not be successfully matched, must be cancelable”?

This is what I found, and it doesn’t support the stricter interpretation.

"If a send is marked for cancellation, then it must be the case that either the send completes normally, in which case the message sent was received at the destination process, or that the send is successfully cancelled, in which case no part of the message was received at the destination. Then, any matching receive has to be satisfied by another send."