Subject: Re: RelCache (aka ELF prebinding) news
To: None <tech-kern@netbsd.org, tech-userlevel@netbsd.org>
From: Ignatios Souvatzis <ignatios@theory.cs.uni-bonn.de>
List: tech-userlevel
Date: 12/03/2002 11:26:02
--SO98HVl1bnMOfKZd
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

Hi,

On Tue, Dec 03, 2002 at 02:27:02AM -0500, Thor Lancelot Simon wrote:

> However, there are significantly better 32-bit CRC polynomials than the
> usual one; the one used by iSCSI is particularly good, against the kind
> of errors seen in data communications links at least.  Whether it is
> better for this purpose, I couldn't really say, because I cannot answer
> even the weak question "given two random bitstrings of length N, what is
> the probability that they will have the same 32-bit CRC result with
> polynomial M", much less "given two ELF object files of length N, what is
> the probability that they will have the same 32-bit CRC result with
> polynomial M", which is what we care about (since the relcache can
> obviously use the file size to eliminate all other collisions).

Not necessarily all. I've not looked closely at ELF binaries, but m68k a.out
binaries for trivial programs tend to be 16kBytes or 24 kiloBytes long...

Regards,
	-is


--SO98HVl1bnMOfKZd
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: 2.6.i

iQEVAgUBPeyGsTCn4om+4LhpAQGNKAgAi64BXm2iESKdJkN4sczU3A/90WpSLd3V
lvgH/DxpOEIyLzNIuNM3rK+ieQtP7nIJXbQuLla5IjQXTTeR0MaOv/wIOM5Xaa4C
lBF3Jf8MgKFujouE66TZ/IZP7dkZpwvbulrx7wN5PEKAr04hsnX4z1lep9MdMZcq
CYC3hNszUt0/DjWNwoN66TkIF/ZxwVBCfvvK93tmD1ovnTZMcRHOhuzmXrp/NIG3
MEUY5DiHIh9ZQQPc181+GW8DOZ5VwKUU+IDcCrUdRIrdLbK4eUO0MVYFSZe7VAD/
7m5MEtUCZMX8ugyQvJqBvypTlIZHyua1sJWuY4OtLAn/i2f5xfop6w==
=+Uat
-----END PGP SIGNATURE-----

--SO98HVl1bnMOfKZd--