Re: [PATCH] 2.5-rmap: VM strict overcommit

[Alan Cox]

On Fri, 2002-07-26 at 11:31, Pavel Machek wrote:
> In what scenario can "strict overcommit" kill?

When the kernel grabs over 50% of RAM. Remember that includes page
tables. I've seen the kernel taking 35% of RAM.

> 
> > +4	-	(NEW) paranoid overcommit. The total address space commit
> > +		for the system is not permitted to exceed swap. The machine
> > +		will never kill a process accessing pages it has mapped
> > +		except due to a bug (ie report it!).
> 
> ...and why is that scenario impossible on "paranoid overcommit"?

You are guaranteed user pages are either discardable or all fit in swap.
If you run out of memory the kernel itself runs out of allocation space
for any kernel owned page.