Re: Extensible hashing and RCU

[Eric Dumazet <dada1@cosmosbay.com>]

On Tuesday 20 February 2007 16:59, Evgeniy Polyakov wrote:
> On Tue, Feb 20, 2007 at 07:49:11AM -0800, Michael K. Edwards 
(medwards.linux@gmail.com) wrote:
> > On 2/20/07, Evgeniy Polyakov <johnpol@2ka.mipt.ru> wrote:
> > >Jenkins _does_ have them, I showed tests half a year ago and in this
> > >thread too. Actually _any_ hash has them it is just a matter of time
> > >to find one.
> >
> > I think you misunderstood me.  If you are trying to DoS me from
> > outside with a hash collision attack, you are trying to feed me
> > packets that fall into the same hash bucket.  The Jenkins hash does
> > not have to be artifact-free, and does not have to be
> > cryptographically strong.  It just has to do a passable job of mixing
> > a random salt into the tuple, so you don't know which string of
> > packets to feed me in order to fill one (or a few) of my buckets.
> > XORing salt into a folded tuple doesn't help; it just permutes the
> > buckets.
>
> Adding XOR with constant value does not change distribution.
> Variable salt will end up with differnet buckets for the same flow.
> It is forbidden - it is not the situation created for passwd/des decades
> ago.

Adding a random hint to jhash (random value picked at boot time, not known by 
attacker) permits to have a secure hash table : An attacker cannot build an 
attack to fill one particular hash chain.

See net/ipv4/route.c (function rt_hash_code()) to see how its used for route 
cache.