[openib-general] mthca FMR correctness (and memory windows)

[Sean Hefty]

Talpey, Thomas wrote:
> This, by the way, is why the fencing behavior of memory windows is so
> painful. I would much rather take an interrupt on bind completion than
> fence the entire send queue. But there isn't a standard way to do that,
> even in iWARP. Sigh.

Can you provide more details on this statement?  When are you fencing the send 
queue when using memory windows?

- Sean