On Wed, 27 Oct 2004 10:08:45 -0400 Hal Rosenstock <[EMAIL PROTECTED]> wrote:
> On Tue, 2004-10-26 at 12:40, Sean Hefty wrote: > > Currently, a call to ib_free_recv_mad does not dereference the mad_agent that > > the mad was given to. The call itself does not access the mad_agent, > > but should a reference be held on the mad_agent while it has a received MAD? > > Looking at the implementation, it appears that a mad_agent could deregister > > with the access layer, then call ib_free_recv_mad, which accesses the > > ib_mad_cache. > > ib_mad_cache is in existence from the time of module insertion to > removal. Deregistering the mad_agent has no effect on its presence so I > don't think the order of ib_free_mad_recv and ib_unregister_mad_agent > matters. I'm just wondering more about whether we should permit an agent to unregister while it has received MADs outstanding. Or, if it makes more sense to block unregister until all MADs have been returned. - Sean _______________________________________________ openib-general mailing list [EMAIL PROTECTED] http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
