netdev - Re: [PATCH] *icsk_ca_priv: Reduce inet_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 for Android: free password hash cracker in your pocket

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

Message-ID: <20080831202238.GA12624@2ka.mipt.ru>
Date:	Mon, 1 Sep 2008 00:22:39 +0400
From:	Evgeniy Polyakov <johnpol@....mipt.ru>
To:	Daniele Lacamera <root@...ielinux.net>
Cc:	netdev@...r.kernel.org, angelo.tornese@...il.com
Subject: Re: [PATCH] *icsk_ca_priv: Reduce inet_connection_sock size?

Hi Daniele.

On Sun, Aug 31, 2008 at 09:19:38PM +0200, Daniele Lacamera (root@...ielinux.net) wrote:
> fast path? I don't think so... the modules allocate the ca_priv once per 
> connection, at the beginning of the connection and only if that specific 
> module needs that.
> Am I missing something?

That's actually a very fast path: connection establishment both for
listened and connection sockets (tcp_init_congestion_control() is called
for TCP_SYN_RECV state too), so you forces every socket allocation (if
appropriate congestion contttrol requires that) to be doubled.

This actually may be not a big problem, since we allocate and free
objects all the time, but still it is additional overhead which can be
avoided. Did you run some perfromance analysis to check if influence is
small or not visible at all?

-- 
	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