Roland Dreier wrote:
> > p = &wqe->flit[1];
> I think flit[] should be __be64.
actually would it make more sense to add a t3_fastreg_pbl_wr or
something like that to the union for the continuation into a second WQE?
- R.
or t3_pbl_fragment
But flit[] should really still be __be64.
Lemme resolve these issues and build with C=1 and CF=-D__CHECK_INDIAN__
and I'll repost.
Thanks,
Stevo
_______________________________________________
general mailing list
[email protected]
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
