[EMAIL PROTECTED] wrote:
> Steve Wise wrote:
>> By the way, this listen backlog for an iwarp rnic is actually the
>> backlog of outstanding TCP connection requests as opposed to iwarp
>> connections.
> 
> Giving this more thought, I think there are at least two
> separate issues.  One is defining what backlog means from the
> perspective of the rdma_listen() API.
> The userspace and kernel APIs don't doesn't necessarily need to have
> the same definition. 
> 
> The second issue that I see is mapping backlog to underlying
> CMs.  A single rdma_listen call can result in separate listen calls
> to underlying devices. (Obviously, a simple mapping is to just pass
> the value down.) 
> 
Agreed, although I would think of it in terms of how many
connection requests the ULP is willing to accept and then
secondly as to what resources the CM/driver/device needs.

_______________________________________________
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