[rds-devel] IB/iWARP code separation

[Steve Wise]

Or Gerlitz wrote:
> On Thu, Nov 6, 2008 at 9:01 PM, Steve Wise <swise at opengridcomputing.com> wrote:
>
>   
>>> Let me see that I am with you: if the completion only indicates that
>>> the operation has been accepted for processing, then what can serve
>>> the rdma initiator as an indication that they can reclaim the buffer?
>>>       
>
>   
>> Completion of any WR means the associated buffers can be reclaimed.
>>     
>
> So Rick's thought about the impossibility to reclaim the buffer at the
> side initiating the rdma following the completion was wrong, I see.
>   

He might have been talking about more than just reclaiming the buffer. 

> What about the other claim saying that errors occurring on the remote
> side of rdma op (e.g invalid remote stag) would not be reported
> through the rdma completion status, is this correct? and if it does,
> how are these errors delivered?
>   

Async events. 

In the iWARP RDMAP protocol, any ingress errors (like bad stag on an 
incoming write or read response) result in a TERMINATE message sent back 
to the peer (and then the connection torn down).  The peer who gets the 
TERMINATE message will map that to an async event.  Looks like cxgb3 and 
amso1100 map it to IB_EVENT_QP_REQ_ERR, and nes maps it to 
IB_EVENT_QP_FATAL.

Steve.