Sean Hefty wrote:
I still need to consider this in more detail to see if there isn't some simpler
solution that we're overlooking.

I was thinking about this more, and I'd like to make sure that we identify all separate issues.

Ignoring RMPP completely, I'm not convinced that we need to take any action. A request will either generate a response or not. If no response is generated, checking that a send is a duplicate provides very little protection. There's only a small window on the send side that such a check would even work, and the receiving side still needs to handle this. And if a response were generated, I don't see that there's any real issue.

I'd like to get some agreement on whether we really need to take any action for non-RMPP MADs, then consider what issues RMPP adds.

- Sean
_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general

To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general

Reply via email to