Subject: Re: Algorithm number/name allocation for opencrypto?
To: None <tech-crypto@netbsd.org>
From: Miles Nordin <carton@Ivy.NET>
List: tech-crypto
Date: 12/08/2007 11:29:10
--pgp-sign-Multipart_Sat_Dec__8_11:28:59_2007-1
Content-Type: text/plain; charset=US-ASCII

>>>>> "tls" == Thor Lancelot Simon <tls@rek.tjls.com> writes:

   tls> hardware that can directly perform SSL record encryption and
   tls> decryption with hash checking.

   tls> what other work should I look at?

solaris has all kinds of crazy web proxying inside kernel space.  I
don't understand it well, but you might want to have a look.  I would
be surprised if I learned it's part of the open source piece, but at
least you could steal some of their metaphors.

Also I noticed some of their crypto accelerator Xnnnnn options for
older boxes like V120 are sold in the form of Ethernet cards.  It's an
interesting world because the hardware was so costly per CPU power
that the i386 ``just throw more CPU's at it'' approach comes with this
HUGE marginal cost you can deflect into complicated crypto cards.
When they architected their crypto hardware/software, they had control
of the whole system, and maybe they had a lot of money to play with
too and did a good job.  or...not.  I dunno.

--pgp-sign-Multipart_Sat_Dec__8_11:28:59_2007-1
Content-Type: application/pgp-signature
Content-Transfer-Encoding: 7bit

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (NetBSD)

iQCVAwUAR1rGVYnCBbTaW/4dAQJJfAQArE3wzp8NvKf4jxlTmNkimVQxZEyjs48D
gtS8GOdtpGoQnDP1Wc+HwjrBEyz9sLBMkJ7+POTJRASDC0r9Cm+9WZuEnbfOuocL
FeKNWN7JjHkRGtd46vZXtz4SHGIJbW3l1rJ9vhuBu68M/evB1L9nMJkU4SDDBnOr
5Kg111U0v5E=
=fePr
-----END PGP SIGNATURE-----

--pgp-sign-Multipart_Sat_Dec__8_11:28:59_2007-1--