netdev - Re: [PATCH net] {xfrm, sctp} Stick to software crc32 even if hardware is capable of that

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: <5257A28B.8000507@windriver.com>
Date:	Fri, 11 Oct 2013 15:02:35 +0800
From:	Fan Du <fan.du@...driver.com>
To:	Neil Horman <nhorman@...driver.com>, <steffen.klassert@...unet.com>
CC:	<vyasevich@...il.com>, <davem@...emloft.net>,
	<netdev@...r.kernel.org>
Subject: Re: [PATCH net] {xfrm, sctp} Stick to software crc32 even if hardware
 is capable of that

On 2013年10月10日 21:11, Neil Horman wrote:
> On Thu, Oct 10, 2013 at 01:51:36PM +0800, Fan Du wrote:
>> igb/ixgbe have hardware sctp checksum support, when this feature is enabled
>> and also IPsec is armed to protect sctp traffic, ugly things happened as
>> xfrm_output checks CHECKSUM_PARTIAL to do check sum operation(sum every thing
>> up and pack the 16bits result in the checksum field). The result is fail
>> establishment of sctp communication.
>>
> Shouldn't this be fixed in the xfrm code then?  E.g. check the device features
> for SCTP checksum offloading and and skip the checksum during xfrm output if its
> available?

sigh... same as my first thought.
However why should xfrm_output check whether the skb is a SCTP one as well as whether
the associated dev is able to do hw SCTP crc32-c checksum in the first place, and then
call SCTP crc checksum value API to write the correct checksum *after* this skb has
leaven SCTP layer???

The checksum operation in xfrm_ouput fits TCP/UDP/IP layer checksum semantics, e.g.
calculate 16bits value by sum almost everything up. Make a SCTP specific exception
in this common path sound wired, as the fix can be done in SCTP codes easily with
minimum changes, so not any bit of consequence for non-SCTP traffic.

And If you/Vlad insist to do so as well as no objection from Steffen/David, I'm
happy to send this revised version as your suggested.

Anyway, I should have separated this patch for two which makes the issues more clear
for you and Vlad to review.

> Or am I missing something?
> Neil
>
>

-- 
浮沉随浪只记今朝笑

--fan
--
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