[ofa-general] [PATCH RFC] RDMA/cxgb3: Fixes for zero-stag.

[Roland Dreier]

 > NFSRDMA was planning on using the zero stag for recvs on iwarp
 > transports.  This boils down to only needing the zero stag + fastreg
 > mrs for NFSRDMA server over T3, avoiding dma_mrs which have limits in
 > the T3 HW.

OK, that makes sense.

 > Hmm.  A new API then?  ib_get_local_lkey() or something?  If IB
 > mandates the lkey be value of 0, then we need don't need to do
 > anything...

No, IB doesn't mandate that.

I don't know if we want a new API or just a new field in struct
ib_device.  We already have the ZERO_STAG capability flag to show when
this makes sense (and we could rename it to be more transport neutral),
and we could just add a local_lkey field, since the value is fixed at
device initialization time.

 - R.