There is debugging code enabled for the loopback interface
in 2.5.x to test out some new features in the networking stack.
This is another reason for the performance drop.
-pavan
> -----Original Message-----
> From: Ralf Baechle [mailto:ralf@linux-mips.org]
> Sent: Thursday, November 07, 2002 8:33 AM
> To: Nagel, Michael
> Cc: lmbench-users@bitmover.com
> Subject: Re: [Lmbench-users] Latest kernel LMBench
> performance results.(wi th HZ=1000)
>
>
> On Wed, Nov 06, 2002 at 10:05:17AM +0100, Nagel, Michael wrote:
>
> > Reading the results I saw a noticeable decrease of the TCP
> performance
> > between the 2.4.x and the 2.5.x kernels. Any explanations
> on that? How
> > about the impact on performance for server-related applications?
>
> That's a typical observation for development kernels and
> should not be valued to highly. In the past month large
> parts of the kernel code have be rewritten. All these
> subsystems and their frequently very subtle interaction with
> each other still need alot of fine tuning which will happen
> as we progress further to a 2.6 release and probably even
> beyond 2.6.0.
>
> Ralf
> _______________________________________________
> Lmbench-users mailing list
> Lmbench-users@bitmover.com
> http://bitmover.com/mailman/listinfo/lmbench-> users
>
This archive was generated by hypermail 2b29 : Thu Nov 07 2002 - 22:00:46 EST