linux-kernel - Re: [PATCH 2/8] firmware: Sigma: Skip header during CRC generation

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] [day] [month] [year] [list]

Date:	Tue, 29 Nov 2011 00:11:38 -0500
From:	Mike Frysinger <vapier@...too.org>
To:	"Lars-Peter Clausen" <lars@...afoo.de>
Cc:	Mark Brown <broonie@...nsource.wolfsonmicro.com>,
	Liam Girdwood <lrg@...com>,
	Andrew Morton <akpm@...ux-foundation.org>,
	linux-kernel@...r.kernel.org, alsa-devel@...a-project.org,
	drivers@...log.com, stable@...nel.org
Subject: Re: [PATCH 2/8] firmware: Sigma: Skip header during CRC generation

On Monday 28 November 2011 02:56:33 Lars-Peter Clausen wrote:
> On 11/25/2011 09:00 PM, Mike Frysinger wrote:
> > On Friday 25 November 2011 03:55:42 Lars-Peter Clausen wrote:
> >> On 11/24/2011 06:21 PM, Mike Frysinger wrote:
> >>> On Thursday 24 November 2011 07:48:21 Lars-Peter Clausen wrote:
> >>>> The firmware header is not part of the CRC, so skip it. Otherwise the
> >>>> firmware will be rejected due to non-matching CRCs.
> >>> 
> >>> that's because you didn't compare to the right value ;).  include the
> >>> CRC -> compare to 0.  omit the CRC -> compare to the CRC value.
> >> 
> >> Does this really work if the CRC is inserted somewhere in the middle of
> >> the bytestream?
> > 
> > i don't think the position matters to the CRC algorithm used by sigmadsp.
> > math principle: a ^ b ^ c is the same thing as b ^ a ^ c and c ^ b ^ a.
> 
> If CRC algorithms  were commutative they would be pretty weak, I guess ;)

that doesn't stop people from implementing them this way ;).  the last CRC 
code i did actually implement and test related to the Blackfin arch did have 
this property (the bootrom).  so i just included the computed CRC value and 
made sure the end result is 0.

> > i could be wrong as to the CRC algo used though.  simple enough for you
> > to check -- i implemented this firmware code based on a spec i wrote up
> > for the sigmadsp peeps; i never actually had real firmware to test with.
> 
> It does not work.

ok, i withdraw my complaints here then :)
-mike

Download attachment "signature.asc " of type "application/pgp-signature" (837 bytes)