netdev - RE: [PATCH net v4 1/1] openvswitch: Fix setting ipv6 fields causing hw csum failure

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: <52b4dbf5-8bd1-d9c7-f672-e7405a91b050@nvidia.com>
Date:   Mon, 21 Feb 2022 11:50:22 +0200
From:   Paul Blakey <paulb@...dia.com>
To:     David Laight <David.Laight@...LAB.COM>
CC:     "dev@...nvswitch.org" <dev@...nvswitch.org>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        Jamal Hadi Salim <jhs@...atatu.com>,
        "Pravin B Shelar" <pshelar@....org>,
        "davem@...emloft.net" <davem@...emloft.net>,
        "Jakub Kicinski" <kuba@...nel.org>,
        Eelco Chaudron <echaudro@...hat.com>,
        Oz Shlomo <ozsh@...dia.com>, Vlad Buslov <vladbu@...dia.com>,
        Roi Dayan <roid@...dia.com>,
        Ariel Levkovich <lariel@...dia.com>
Subject: RE: [PATCH net v4 1/1] openvswitch: Fix setting ipv6 fields causing
 hw csum failure




On Sun, 20 Feb 2022, David Laight wrote:

> From: Paul Blakey
> > Sent: 20 February 2022 13:21
> > 
> > Ipv6 ttl, label and tos fields are modified without first
> > pulling/pushing the ipv6 header, which would have updated
> > the hw csum (if available). This might cause csum validation
> > when sending the packet to the stack, as can be seen in
> > the trace below.
> > 
> > Fix this by updating skb->csum if available.
> ...
> > +static inline __wsum
> > +csum_block_replace(__wsum csum, __wsum old, __wsum new, int offset)
> > +{
> > +	return csum_block_add(csum_block_sub(csum, old, offset), new, offset);
> > +}
> 
> That look computationally OTT for sub 32bit adjustments.
> 
> It ought to be enough to do:
> 	return csum_add(old_csum, 0xf0000fff + new - old);
> 
> Although it will need 'tweaking' for odd aligned 24bit values.
> 
> 	David
> 
> -
> Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
> Registration No: 1397386 (Wales)
> 
> 

I'm not comfortable with manually implementing some low-level
bit fideling checksum calculation, while covering all cases.
And as you said it also doesnt have the offset param (odd aligned).
The same could be said about csum_block_add()/sub(), replace()
just uses them. I think the csum operations are more readable and
should  be inlined/optimize, so can we do it the readable way,
and then,  if needed, optimize it?