Re: MMX based IP-checksumming patch, 2.1.105, RFC
Alexander Kjeldaas (astor@guardian.no)
Sat, 13 Jun 1998 20:11:47 +0200
On Sat, Jun 13, 1998 at 04:52:58AM -0700, mingo@hal.cobaltmicro.com wrote:
>
> the core unrolled loop of the csum_partial() function does cold-cache
> checksumming at a speed of 270 MB/sec on my PII box (stock function does
> 210 MB/sec), in-cache performance is 540 MB/sec vs. 350 MB/sec, but for
> most RL MTU's the FPU save/restore operation eats a considerable amount of
> the saved cycles ...
Could you test them against the PPro routines in Davem's tree? They
should be much faster than the ones in the vanilla kernel.
astor
--
Alexander Kjeldaas, Guardian Networks AS, Trondheim, Norway
http://www.guardian.no/
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu