netdev - Re: [RFC PATCH] ethtool: add IPv6 to the NFC API

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: <56A260D9.3020004@solarflare.com>
Date:	Fri, 22 Jan 2016 17:03:21 +0000
From:	Edward Cree <ecree@...arflare.com>
To:	Alexander Duyck <alexander.duyck@...il.com>
CC:	Ben Hutchings <ben@...adent.org.uk>,
	Netdev <netdev@...r.kernel.org>
Subject: Re: [RFC PATCH] ethtool: add IPv6 to the NFC API

On 21/01/16 22:48, Alexander Duyck wrote:
> On Thu, Jan 21, 2016 at 11:14 AM, Edward Cree <ecree@...arflare.com> wrote:
>> +       __be32  spi;
> Is this supposed to be the flow label, or is this the IPSec security
> parameter index?  If it is the flow label you may want to rename it.
> If it is supposed to be the IPSec security parameter index this might
> belong in a different flow definition since it is not actually a part
> of the IPv6 header.
It's the IPSec SPI; I just blindly copied what ethtool_ah_espip4_spec had.
I guess splitting out three different spec structs as per ipv4 would make
this somewhat clearer.
Would the flow label be useful thing to include?  I would have thought it'd
be a bit too short-lived normally.
>> +       __u8    tos;
>> +       __u8    proto;
>> +};
>> +
> Technically the name of the field for proto is nexthdr.
But will NICs filter on the actual nexthdr, or on the *last* nexthdr in the
chain of IPv6 options?  If the latter, calling it nexthdr might be misleading.
(I've just checked what sfc will do, it filters on the last nexthdr.)

Will spin a v2 shortly.

-Ed