netdev - Re: [RFC 1/3] tcp: randomize tcp timestamp offsets for each connection

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] [day] [month] [year] [list]

Message-ID: <1472178854.14381.180.camel@edumazet-glaptop3.roam.corp.google.com>
Date:   Thu, 25 Aug 2016 19:34:14 -0700
From:   Eric Dumazet <eric.dumazet@...il.com>
To:     Florian Westphal <fw@...len.de>
Cc:     netdev@...r.kernel.org
Subject: Re: [RFC 1/3] tcp: randomize tcp timestamp offsets for each
 connection

On Fri, 2016-08-26 at 01:46 +0200, Florian Westphal wrote:

> Wrt. making randomization optional:
> 
> Would you go for another sysctl or should I just change
> secure_tcpvX_sequence_number to check for tcp_timestamps == 2 mode?
> 
> 	*tsoff  = sysctl_tcp_timestamps == 2 ? hash[1] : 0;

I've not yet look at your "add tcp_timestamps=2 mode to force tsecr
validation on ofo segments" patch.
It looks a bit scary to me :(

I would rather have a separate sysctl, or maybe a per route attribute.

Note that this could be done later.

> 
> Could also use a static key but I don't think its worth it vs. md5 cost.
> What do you think?

A static key wont help a lot I think ;)