netdev - Re: csum offload and af

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: <20080813010935.GA29815@gondor.apana.org.au>
Date:	Wed, 13 Aug 2008 11:09:35 +1000
From:	Herbert Xu <herbert@...dor.apana.org.au>
To:	David Miller <davem@...emloft.net>
Cc:	netdev@...eo.de, rusty@...tcorp.com.au, anthony@...emonkey.ws,
	netdev@...r.kernel.org
Subject: Re: csum offload and af_packet

On Tue, Aug 12, 2008 at 05:55:32PM -0700, David Miller wrote:
> 
> The whole point of the auxdata thing was that if your tool is smart
> you can get the notification that the checksum is going to be HW
> computed and therefore keep capturing at a very low cost.
> 
> But for a stupid AF_PACKET user, like dhclient currently is, there is
> no harm in COW'ing the packet header area and calculating the checksum
> for them in this case.

Right I'm not against adding this to AF_PACKET.

It's just that the virtualisation problem that started all this
is just as easily resolved by user-space.  In fact the dhcp-client
code in RHEL/Fedora has been patched to do exactly this a long time
ago.

> I kind of hope that libpcap/tcpdump has been taught about auxdata
> by now, has it? :-/

Unfortunately no.  This also means that if we do do this in AF_PACKET,
then running a packet dump on a fast outbound interface may become more
expensive.

Cheers,
-- 
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@...dor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
--
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