On Thu, Mar 16, 2006 at 10:52:01PM +0100, Dag Arne Osvik wrote:
> 
> 187.5 is for the existing x86-64 asm module running on an AMD Athlon 64. 
>  None of the above numbers are for my code, since I haven't made a 
> kernel module of it just yet, and hence can't report tcrypt testing 
> results for it.

OK.

> The only real news I'm presenting is that my code performs CBC-like 
> encryption in about 233 cycles on a 64-bit Intel Celeron.  Apart from 
> that, I was very surprised that CBC-mode encryption in the kernel is so 
> slow compared to ECB.  The difference is really just four xor operations 
> per block.

I see.  The good news is that with the upcoming parameterised crypto
stuff you'll be able to have an asm-optimised CBC implementation just
like AES.  It'd be interesting to see whether the speed-up is coming
from doing CBC in asm or just eliminating indirect function calls.

Cheers,
-- 
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <[EMAIL PROTECTED]>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
-
To unsubscribe from this list: send the line "unsubscribe linux-crypto" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to