On 7/9/2015 3:03 AM, Jason Gunthorpe wrote:
On Wed, Jul 08, 2015 at 01:32:05PM -0700, 'Christoph Hellwig' wrote:
On Wed, Jul 08, 2015 at 01:08:42PM -0600, Jason Gunthorpe wrote:
Then, what is left is all remote MRs and maybe it will be clearer what
to do about them then...
From looking at that for a while the APIs needed seem pretty simple
to me from a consumer perspective:
struct rdma_mr *rmda_alloc_mr(struct ib_pd *pd, unsigned int nr_pages);
void rdma_free_mr(struct rdma_mr *mr);
The major trouble with that is that the new MR types work by posting
work to the send queue, that work creates the MR.
I don't know all the details of how those schemes work, but it doesn't
look like it fits into this model ?
Maybe if rdma_register_sg was the only API and it accepted a QP, under
the hood it could re-use a pooled/create a non-queued (F)MR, or issue
FRMR, or work with indirect?
I think it's confusing to possibly do completely different registration
methods which has different implications on the consumer RDMA channel.
It would be nice if it is that simple :)
It can be simple if we consolidate on a single method.
Sagi, are FMRs, FRMRs and Indirect MRs documented any place? I can't
recall off hand if FRMR was ever published by the IBTA?
I don't know if FMRs are in IBTA.
FRMRs are in under 10.6.3.7 FAST REGISTRATION (vol 1)
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
