Ignatios Souvatzis <is@netbsd.org> writes:
> > Yes. SHMMAXPGS should be nuked from orbit -- it should dynamically
> > size itself. (We should also somehow garbage collect segments that are
> > no longer associated with running processes.)
> 
> Uhm - but those could be accessed later by a new process, can't they?

In theory, yes. In practice, they're usually programming errors or
crashes. (We could have a sysctl to turn off the behavior but I doubt
many would use it.)


--
Perry E. Metzger		perry@wasabisystems.com
--
NetBSD: The right OS for your embedded design. http://www.wasabisystems.com/