[PATCH 2/2] Avoid slow integer multiplication and division with blocksize calculations.

Werner Koch wk at gnupg.org
Fri Nov 30 10:21:08 CET 2012

On Thu, 29 Nov 2012 20:09, jussi.kivilinna at mbnet.fi said:

> Well, I currently only have access to x86 machines and there this
> didn't have easily measurable effect.

Okay, I was just curious.

> However if I leave 'buffer xor' patch out, and change cipher-ctr.c
> loop to use '& c->blockmask' instead of '% blocksize', AMD Phenom
> II/x86-64 sees following improvement:




Die Gedanken sind frei.  Ausnahmen regelt ein Bundesgesetz.

More information about the Gcrypt-devel mailing list