[openib-general] [PATCH 1/2] perftest: enhancement to rdma_bw to allow use of RDMA CM

[Pradipta Kumar Banerjee]

Steve Wise wrote:
> On Thu, 2006-07-06 at 09:20 +0300, Michael S. Tsirkin wrote:
>> Quoting r. Steve Wise <swise at opengridcomputing.com>:
>>> For instance, in the bi-directional bandwidth tests, one side could
>>> finish the test and issue an rdma_disconnect() while the other end still
>>> has pending rdma writes.  This is an error condition in iWARP.  So they
>>> sync up at the end with a message exchange before disconnecting.
>>>
>>> Make sense?
>> I see. Is this the only message exchange there is?
> 
> 
> There are 2 message exchanges to comply with IWARP spec: 
> 
> 1) an exchange right after connection setup to force the client to be
> the first side to send an RDMA message.  So the server waits for a RECV
> completion and the client posts a "lets start" SEND.
Steve,
Thanks for clarifying this. However I have a doubt. Wouldn't server doing a 
rdma_listen() and the client doing a rdma_connect() suffice ?
I am going through the iWARP spec to understand why this first exchange is 
required and why a listen/connect sequence is not enough. Any pointers will be 
really helpful.

Thanks,
Pradipta
> 
> 2) an exchange at the end of the test to ensure all RDMA operations are
> done on the QPs.  Both client and server post a "I'm done" SEND and
> await a "I'm done" completion before disconnecting.
> 
> 
> Steve.
> 
>