[openib-general] Re: [PATCH] request/response matching in MAD code

Mon Oct 4 09:38:13 PDT 2004

On Mon, 04 Oct 2004 11:53:21 -0400
Hal Rosenstock <halr at voltaire.com> wrote:

> It is a minor efficiency gain. The downside is the following:
> If the client sets the hi_tid wrong (not corresponding to the mad_agent
> supplied), the response either gets thrown away (if there is no hi_tid
> match) or delivered to the "wrong" client callback.
> 
> A "similar" scenario exists when the client specifies a timeout and the
> send is "unsolicited". In this case, the send although never getting a
> response is timed out and not completed until then. It could be
> completed when the send completion occurs ignoring the timeout.

I don't have strong feelings on this either way.  What I'd like to avoid is pushing too much protocol knowledge into the access layer for a given management class.  Some of it is difficult to avoid and still provide some useful functionality.

As an example, a CM REP MAD is solicited.  The access layer shouldn't modify the TID, but it makes sense for the user to specify a timeout value.

- Sean