[Soekris] net4801 bridging firewalls benchmarks results

Chris Cappuccio chris at nmedia.net
Sun Oct 9 08:17:05 UTC 2005


Jason Dixon [jason at dixongroup.net] wrote:
> 
> P.P.S.  I'm not surprised by the OpenBSD speeds;  I believe the  
> OpenBSD sis driver still has issues.  I have no evidence to back this  
> up, just anecdotal evidence (the maxxed out irq's, for one).
> 

So, what in the world is "maxxed out irq's" ?

Anyways, the OpenBSD driver is very similar to the FreeBSD one, as that's
where it's derived from.  The problem isn't the driver, it's that the OpenBSD
ip forwarding code path isn't quite as optimized for packet forwarding
as FreeBSD has done.  Polling helps as well, but I don't see why using
setting Interrupt Hold-Off register doesn't produce results identical to
polling, given similar parameters.  (If someone could explain what is probably
blatantly obvious to them about the differences, I would appreciate it.)

Although, I agree I've never seen numbers for FreeBSD on Soekris as high as
this before.  (Then again, most of my memory is probably of net45xx performance
and not net48xx.)  But even with polling, the idea that a net4801 uses only
4-8% CPU while pushing nearly full speed is interesting and seems 
inaccurate.

-- 
"Attacks always get better; they never get worse."
  -- "Old NSA saying"



More information about the Soekris-tech mailing list