[openib-general] DAPL for openib
Michael S. Tsirkin
mst at mellanox.co.il
Wed Aug 25 13:44:21 PDT 2004
Hello!
Quoting r. Michael S. Tsirkin (mst at mellanox.co.il) "Re: [openib-general] DAPL for openib":
> Hello!
> Quoting r. Roland Dreier (roland at topspin.com) "Re: [openib-general] DAPL for openib":
> >
> > Sean> We need to discuss how to route MADs to the SMA. Are you
> > Sean> suggesting that the SMA would register for MADs the same as
> > Sean> any other client? My plan was to extend the ib_mad_reg_req
> > Sean> structure as needed. I limited it to version, class, method
> > Sean> currently, since it easily allows O(1) routing. We could
> > Sean> add in AttributeID and AttributeModifier if those are
> > Sean> needed.
> >
> > I don't know how the SMA should be handled. In the Topspin drivers
> > it's a bit of a special case -- it gets to see all MADs before any
> > other consumers. We could do the same, or have the SMA register.
> >
> > - R.
>
> I'm not sure I understand the semantics -
> if SMA simply registers without supplying
> version, class or method, will it get to see all MADs?
>
Sorry, I just noticed this is being addressed already:
From contrib/intel/TODO:
- Need to extend ib_mad_reg_req to support snooping.
could use masks for class, version, methods to snoop all.
could add a registration type - receive MADs vs. view MADs.
registration for view MADs would need to apply to sends as well.
view MAD would solve the SMA issue , right?
MST
More information about the general
mailing list