Re: Interesting scheduling times

Claude Gamache (cgamache@cae.ca)
17 Sep 1998 10:21:24 -0400


Richard Gooch <rgooch@atnf.csiro.au> writes:

> OK:
>
> CPU process switch thread switch Kernel version
> Pentium 100 12 12 2.1.109
> Pentium 100 19 9 2.1.122-pre2
> PPro 180 8 4 2.1.122-pre2
> Pentium/MMX 200 6 2 2.1.122-pre2
> Pentium/MMX 200 4 2 2.1.104
>
> Even more interesting: 2.1.122-pre2 is slower for process switching
> and is faster for thread switching than 2.1.104/2.1.109.
>

Here are the results for a dual PII 400 MHz and a PPro 200 MHz (256 kb
cache):
process thread Kernel
CPU switch switch version
--------------------------------------------------------------
Dual PII 400 MHz,(nice -20) 5 us 5 us 2.0.36-pre8
Dual PII 400 MHz,(nice 0) 10 us 9 us 2.0.36-pre8
PPro 200 MHz(256 kb cache) 12 us 7 us 2.1.122-pre3

With the PPro, the priority does not make any difference. Although,
with the dual PII, it makes a difference !?

We also noted in the 2.1.x kernels (vs kernel 2.0.xx) a general
slowdown in terms of performance with our applications. We posted a
few times regarding this and we have not yet clearly pointed out what
causes these slowdowns.

Various individuals have posted on the kernel mailing list performance
issues with many different pieces of software (Electric Fence, Xserver,
compilation of kernel, etc; and now process switching).

Could part of these slowdowns be linked with the process switching
time of the recent kernels ?

If not, what could cause these slowdowns ?

Regards,

Claude

-- 
  Claude Gamache, CAE Electronique Ltee, 8585 Cote-de-Liesse  
  Saint-Laurent,  Quebec, Canada H4T 1G6                        
  Email: cgamache@cae.ca  Tel.: (514) 341-2000 x3194, Fax: (514) 734-5612

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/