Re: pre9-4 OOM VM lockups

From: Jesse Pollard (
Date: Wed May 24 2000 - 13:28:54 EST

Rik van Riel <>:
> On Wed, 24 May 2000, Jesse Pollard wrote:
> > Stefan Monnier" <monnier+lists/linux/kernel/news/@RUM.CS.YALE.EDU>:
> > >
> > > >>>>> "t" == t n vanderleeuw <> writes:
> > > > The only time when it went wrong in 2.2.16pre3 was when the X server
> > > > got killed and thus the graphics card was left in an unpredictable
> > > > state. In the next test I shut down X first and all was mostly fine.
> > >
> > > So if I understand it correctly, Rik's OOM Killer still hasn't been
> > > merged into the standard kernel ?
> >
> > I thought it had been. It just can't guarantee the
> > identification of the real hog. It is just better with than
> > without.
> It's not in the kernel. At the moment I'm not pushing to get
> it included either since identifying the point _where_ we run
> out of memory is really hard...
> Once VM is rewritten a bit, I'll probably try to have the
> OOM killer included into the kernel.
> About VM: I've written a number of proposals for VM changes to
> the linux-mm mailing list (,
> but seem to never get any reactions on my plans ... this means
> I could start implementing stuff, but maybe I overlooked some
> critical thing. ;((

Just our difference of opinion on user virtual memory quotas -- and I'm
willing to wait for the VM rewrite for that. The patch I got earlier
works quite well for single user systems/dedicated system to prevent
a user from killing the system (the controlable overcommit patch from
a month or so ago).

And thanks for the reference the linux-mm archives. I'm not a subsriber
(I already get 1-200 messages/day).

Jesse I Pollard, II

Any opinions expressed are solely my own.

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
Please read the FAQ at

This archive was generated by hypermail 2b29 : Wed May 31 2000 - 21:00:12 EST