possible mpi-pow improvement

[Werner Koch]

On Fri,  6 Sep 2013 10:41, gniibe at fsij.org said:

> ====================== original =====================
> $ ./tests/benchmark rsa
> Algorithm         generate  100*sign  100*verify
> ------------------------------------------------
> RSA 1024 bit         340ms     860ms        30ms
> RSA 2048 bit         870ms    5510ms       110ms
> RSA 3072 bit        6440ms   16930ms       210ms
> RSA 4096 bit       17470ms   37270ms       360ms


> My possible change:
> ====================== k-ary, MUL instead of SQR =====
> Algorithm         generate  100*sign  100*verify
> ------------------------------------------------
> RSA 1024 bit         280ms     710ms        30ms
> RSA 2048 bit         960ms    4410ms       110ms
> RSA 3072 bit       17680ms   12990ms       220ms
> RSA 4096 bit       12280ms   29550ms       360ms
>
> Any comments are appreciated.

Thus you change is even an improvement for the general case.

Can you please change your patch to conditionally include the k-ary
multiply but enable it right away.


Shalom-Salam,

   Werner

-- 
Die Gedanken sind frei.  Ausnahmen regelt ein Bundesgesetz.