GCJ-friendly RIPEMD-1xx + Khazad. was: [GNU Crypto] GCJ-friendly MD5

[Raif S. Naffah]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: RIPEMD160

hello Casey,

On Friday 06 December 2002 19:09, Casey Marshall wrote:
> Raif S. Naffah wrote:
> | hello there,
> |
> | yesterday, MD5 was at 37.7 MB/s.  today it's at 75.2 MB/s.
>
> Any thoughts about putting this version into the main tree?...

for sure.  thanks for trying it :-)

i'll do that tonight with...

* ripemd128 was 19816.61 KB/s, now is 91609.99 KB/s,
* ripemd160 was 15837.86 KB/s, now is 58337.07 KB/s

and

* khazad was (E/D) 6430.04/6414.20 KB/s, now is 7193.83/7247.22 KB/s

i believe we now have a good portfolio of GCJ-friendly implementations 
of our algorithms.  the performance of RIPEMD, SHA and MD5 are 
exceptionally good.


i'll be spending the next few days looking at how we can use Mauve 
instead of JUnit, at least for the GCJ-friendly build of GNU Crypto to 
closer align the library with GNU Classpath and GCJ.  i'll keep you 
posted.


cheers;
rsn
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)
Comment: Que du magnifique

iD8DBQE98cFR+e1AKnsTRiERA1yKAKC0MIa6b0TwUtMkWX6wFku/48uRpQCfXfWT
ylhCCEVjhZBQntiZfphsC6U=
=SfqX
-----END PGP SIGNATURE-----