Re: new loop aes and speedup

[Jari Ruusu <jari.ruusu@pp.inet.fi>]

Newsmail wrote:
> nope, I meant speed comparaison between loopaes 1.6h and 1.6i for 2.4. so
> apparently there shouldnt be any speed difference right?

Other possible speedups:
1)  If you are using x86 boxes, and your earlier kernel was compiled for 386
    or 486, and you now have it recompiled for Pentium or better x86.
    Pentium and better x86 targets automatically get optimized assembler
    implementation of AES cipher, which is about twice as fast as the C
    implementation.
2)  You are running a kernel with better VM.
3)  Better C compiler code generation.

Just upgrading loop-AES v1.6h to v1.6i should not show better performance.
Last major speed improvement was in December 2001 when optimized AES cipher
implementation for Pentium or better x86 processors was added.

Regards,
Jari Ruusu <jari.ruusu@pp.inet.fi>

-
Linux-crypto:  cryptography in and on the Linux system
Archive:       http://mail.nl.linux.org/linux-crypto/