[openib-general] [PATCH][iWARP] Added provider CM verbs andqueryprovider methods

[Steve Wise]

> -----Original Message-----
> From: Sean Hefty [mailto:sean.hefty at intel.com] 
> Sent: Thursday, August 25, 2005 1:36 PM
> To: 'Steve Wise'; 'Tom Tucker'; openib-general at openib.org
> Subject: RE: [openib-general] [PATCH][iWARP] Added provider 
> CM verbs andqueryprovider methods 
> 
> >The Ammasso 1100 does do 100% connection setup.  That's why we're
> >pushing connection establishment verbs into the device struct.  IMO,
> >these functions are analagous to the process_mad function in the
> >ib_device structs, which has no meaning to an iwarp device.  
> So I think
> >we have to admit up front, that the ib_device struct really has
> >Infiniband-specific verb functions as well as iWARP-specific verb
> >functions, and that's ok.  (or maybe not :-)
> 
> Does it do the connection setup in hardware or software?
> 

Hardware (actually firmware running on the amso1100 adapter).


> I agree that there are IB specific functions in the IB device 
> structure today,
> but I'm not sure of the best way to define an RDMA device 
> structure moving
> forward.  Longer term, it may make sense to separate out the 
> transport specific
> functions, in which case we can start moving in that direction.
> 
> 
> >Assuming each RNIC supported some raw way to send and 
> receive ethernet
> >frames, then you could implement TCP, IP, ICMP, ARP etc al 
> as a common
> >stack to setup connections.  I don't think we want to do this?
> 
> Are you saying that your RNIC cannot send and receive raw 
> ethernet frames and
> act as a plain ethernet NIC?  I was assuming that this was a 
> given for any RNIC
> device.

It can, but that uses a different MAC address.