[openib-general] basic IB doubt

[Sean Hefty]

>Actually, if a hardware implementation provided the same performance
>(in this case latency) by polling on a CQ as one where polling on
>memory was guaranteed to work, the customer may actually prefer the
>"standard" implementation.

Polling on a CQ involves a function call, synchronization to the CQ, and
formatting a structure to return to the user.  I don't see this ever being
faster than polling memory.

If the data is received in order, there's no additional delay between writing
the data and polling on memory.  A CQ entry requires a separate write, plus the
overhead mentioned above.  Even if the data arrived out of order, only the last
byte needs to be deferred.  Writing that byte shouldn't be any slower than
writing a CQ entry.

- Sean