Let the flames begin... [was Re: [patch 3/9] mempool - Make mempools NUMA aware]

[Pavel Machek <pavel@suse.cz>]

On So 28-01-06 23:59:07, Pavel Machek wrote:
> Hi!
> 
> > >If sending routines can work with constant ammount of memory, why use
> > >kmalloc at all? Anyway I thought we were talking receiving side
> > >earlier in the thread.
> > >
> > >Ouch and wait a moment. You claim that GFP_KERNEL allocations can't
> > >block/sleep? Of course they can, that's why they are GFP_KERNEL and
> > >not GFP_ATOMIC.
> > >  
> > I didn't meant GFP_KERNEL allocations cannot block/sleep? When in 
> > emergency, we
> > want even the GFP_KERNEL allocations that are made by critical sockets 
> > not to block/sleep.
> > So my original critical sockets patches changes the gfp flag passed to 
> > these allocation requests
> > to GFP_KERNEL|GFP_CRITICAL.

(I'd say Al Viro mode, but Al could take that personally)

IOW: You keep pushing complex and known-broken solution for problem
that does not exist.

(Now you should be angry enough, and please explain why I'm wrong in
easy to understand terms, so that even I will understand that we need
critical sockets for kernels in emergency).
								Pavel
-- 
Thanks, Sharp!