Messages in this thread | | | Date | Fri, 15 May 2015 17:02:57 +0200 | From | Michael Wang <> | Subject | Re: [PATCH RFC] Documentation/infiniband: Add docs for rdma-helpers |
| |
On 05/15/2015 04:40 PM, Doug Ledford wrote: [snip] > > The test itself doesn't mean that. It means we need a RoCE address > (it's true when transport is IB and link layer is Ethernet). That we > *use* it during connectionless communication because we have to generate > our own address vector for the packet while during connected queue pair > use the address vector is created by the card using the queue pair > information is just the circumstance of its use. And even though a > disconnected queue pair isn't solidly connected to a remote endpoint, it > is solidly bound to an adapter that requires either an IB or Ethernet > address family. Maybe this to resolve your issue with the wording:
Thanks for the explain :-) The term 'connectionless' still sounds a little strange to me when it's just means no HW support on creating address vector, but I can understand the concept.
> > This helper is true when the address family of this queue pair is of the > Ethernet (RoCE) variety.
Sounds good, will be merged in next version :-)
Regards, Michael Wang
> >
| |