[openib-general] Re: Re: [PATCH] RFC Verbs: add support for transport specific verbs

[Roland Dreier]

    Sean> After we decide on which approach to use, we can determine
    Sean> if other calls should be separated out, such as query_pkey.
    Sean> Roland/Hal, do you have a particular preference?

I think this is all a bit of a bicycle shed discussion.  As long as
there isn't a runtime penalty, I don't see much of a reason to care.

Rather than making these small changes piecemeal, I would much rather
see a complete patch series that brings a working and relatively clean
Ammasso driver to the trunk.  Then we can merge to the upstream kernel
soon, rather than having OpenIB svn drift further and further from
upstream.

On a (related) tangent, are the RMPP changes ready to merge upstream?
Can someone prepare a patch series to do the merge?  The MAD code also
looks like it has diverged quite a bit, and it would be good to get
upstream back in sync.

 - R.