On Tue, 2005-03-15 at 14:44 -0800, Chris Wright wrote: > Luckily, your kmalloc(-1) example would fail (there's no slab that > big), The failure mode will be a BUG(). -- dwmw2