[PATCH]: fdalloc optimization

Andre Oppermann andre at freebsd.org
Mon Aug 22 11:12:25 GMT 2005


Poul-Henning Kamp wrote:
> In message <20050822083320.GA32668 at stud.fit.vutbr.cz>, Divacky Roman writes:
> 
> 
>>   N           Min           Max        Median           Avg        Stddev
>>x 100         39957        959570         41444      87169.69     201011.84
>>+ 100         37786        962107         38766      66353.09     158258.74
>>No difference proven at 95.0% confidence
>>
>>anyway - commenting out duplicate code MUST speed it up, maybe insignificantly
>>but still ;)
> 
> 
> It used to be that life was that simple, but with three levels of caches and
> deep pipelines it isn't anymore.
> 
> But taken from the other side: you can also read this as your patch does no
> damage to performance, so if code clarity is improved (havn't looked at your
> patch myself) then that would be another good reason.

Danfe posted about some DFBSD work in that area an hour ago.

-- 
Andre


More information about the freebsd-current mailing list