Something Seriously Wrong, Kernel memory death.

Linus Torvalds (
Fri, 9 May 1997 23:12:56 -0700 (PDT)

Isnt there something seriously wrong with a system in which its faster to
create a process than to kill it?

And of course, I thought that the "most recently created" process was the
one killed anyways to free up memory. Isnt it?


On Wed, 7 May 1997 wrote:
> I think it would almost be dangerous to kill off randium processes. It
> would make more sense to kill off the newest processes untill there is
> enough memory. That way, as in the above example, you wouldn't kill the x