netdev - Re: Extensible hashing and RCU

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20070220175550.GB26961@2ka.mipt.ru>
Date:	Tue, 20 Feb 2007 20:55:50 +0300
From:	Evgeniy Polyakov <johnpol@....mipt.ru>
To:	Eric Dumazet <dada1@...mosbay.com>
Cc:	"Michael K. Edwards" <medwards.linux@...il.com>,
	David Miller <davem@...emloft.net>, akepner@....com,
	linux@...izon.com, netdev@...r.kernel.org, bcrl@...ck.org
Subject: Re: Extensible hashing and RCU

On Tue, Feb 20, 2007 at 06:20:26PM +0100, Eric Dumazet (dada1@...mosbay.com) wrote:
> > Hmm, I've just ran following test:
> > 1. created 2^20 hash table.
> > 2. ran in loop (100*(2^20) iterations) following hashes:
> >  a. xor hash (const_ip, const_ip, random_word)
> 
> So what ? to attack me you want to send 100*2^20 packets every minute ?

:) No, I will specially craft 1000 packets which will hist the same
chain.

> Thats nonsense... If you really can send so many packets, My pipe is full 
> whatever I do of received packets. No Algo will protect me, even designed by 
> Einstein.

Did you ever read what I wrote?
It is test, which shows that 
1. jenkins has problems
2. it is two times slower than xor

How to explot problem in a real world is out of that research, but it is
enough to say that it is broken.

> If you look again at route cache, you will see chains length are limited by 
> elasticity factor, that is usually 8... No need to try to reach 100 entries 
> in a chain.
> 
> Yes, I can destroy Russia sending 2^10 nuclear weapons on major cities. You 
> really should build a bunker right now :)

France only has 100 delivery vehicles (about 50 submarines and 50
Mirages) - so no, I will not :)

> Now try to build an attack with 100 packets per second... and I will try to be 
> smart too.

Depending on the end result... Wanna buy me (or suggest) couple of bottles of 
good not expensive french wine? :)

Here is a dump of possible addr/port pairs which end up badly
distributed:

8e363a50:27652 -> c0a80001:20480
8e363a50:35529 -> c0a80001:20480
8e363a50:40919 -> c0a80001:20480
8e363a50:46720 -> c0a80001:20480

they produce the same hash value in the test described above.

-- 
	Evgeniy Polyakov
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html