[openib-general] DMA mapping abuses in MAD layer

[Sean Hefty]

Roland Dreier wrote:
> We probably still want to handle gather lists for posting sends I
> think.  Another (rather unrelated) issue that I just noticed the other
> day is that something like sending a response to a GetTable request
> for PortInfo for every port in a large fabric is going to end up
> sending a very large RMPP message, probably too large to fit in a
> single kmalloc()ed buffer.  So I don't think we should require that
> all send requests have a single gather entry.

We can change the ib_mad_send_buf to allow chaining them together.

The single SGE restriction was just a limitation of the initial implementation. 
  Supporting an arbitrary breaking of a MAD buffer across multiple SGEs is 
difficult, but if we can control the SGE sizes, this should be doable.  I think 
using ib_mad_send_buf in post_send_mad makes supporting this easier.

Does anyone else have any other ideas on how to fix this issue?

- Sean