[ On Wed, April 15, 1998 at 09:29:35 (-0700), Eduardo E. Horvath wrote: ]
> Subject: Re: Real vfork() (was: third results) 
>
> One important issue all of the COW proponents are ignoring is swap space
> overcommit.  This issue was hashed out in comp.unix.internals a few months
> ago.  [....]

Again?!?!  ;-)

Yes, that's a very important point actually.

BTW, I also find any OS that overcommits swap to be evil.  If fork() is
failed then the application can resort to any number of recovery
procedures, but if it gets an async. signal from the OS saying it's the
next candidate for swap space recovery then it's often stuck between a
rock and a hard place since it is awfully hard for either the
application or the kernel to second guess the user(s) on issues of
priority and importance (or esp. to arbitrate between users!).

-- 
							Greg A. Woods

+1 416 443-1734      VE3TCP      <gwoods@acm.org>      <robohack!woods>
Planix, Inc. <woods@planix.com>; Secrets of the Weird <woods@weird.com>