1998-12-21 10:56:39 +00:00
|
|
|
RIPEMD-160
|
|
|
|
|
http://www.esat.kuleuven.ac.be/~bosselae/ripemd160.html
|
|
|
|
|
|
|
|
|
|
This is my implementation of RIPEMD-160. The pentium assember is a little
|
|
|
|
|
off the pace since I only get 1050 cycles, while the best is 1013.
|
|
|
|
|
I have a few ideas for how to get another 20 or so cycles, but at
|
2009-02-14 21:49:38 +00:00
|
|
|
this point I will not bother right now. I believe the trick will be
|
1998-12-21 10:56:39 +00:00
|
|
|
to remove my 'copy X array onto stack' until inside the RIP1() finctions the
|
|
|
|
|
first time round. To do this I need another register and will only have one
|
|
|
|
|
temporary one. A bit tricky.... I can also cleanup the saving of the 5 words
|
|
|
|
|
after the first half of the calculation. I should read the origional
|
|
|
|
|
value, add then write. Currently I just save the new and read the origioal.
|
|
|
|
|
I then read both at the end. Bad.
|
|
|
|
|
|
|
|
|
|
eric (20-Jan-1998)
|
