Re: ULE/yielding patch for testing.

From: Marc Fonvieille <blackend_at_freebsd.org>
Date: Mon, 8 Oct 2007 16:31:07 +0200
On Mon, Oct 08, 2007 at 04:27:21PM +0200, Marc Fonvieille wrote:
> > 
> > Is this a SMP kernel on a UP machine?  If so can you try without SMP in 
> > your config?
> >
> 
> Without SMP and apic, the number of context switches is lower:
> 
>  procs      memory      page                    disks     faults cpu
>  r b w     avm    fre   flt  re  pi  po    fr  sr ad4 cd0   in   sy   cs us sy id
>  0 0 0  501656  89424     0   0   0   0     0   0   0   0 1266 3532 1511 8  7 86
>  0 0 0  501656  89424     0   0   0   0     0   0   0   0 1252 8127  964 11  6 83
>  1 0 0  501656  89424     0   0   0   0     0   0   0   0 1253 1514  973 2  7 92
>  3 0 0  501656  89424  2647   0   0   0  2662   0   6   0 1397 12460 2663 42 24 34
>  2 0 0  501656  89424   379   0   0   0   375   0   0   0 1307 16338 2528 81 17  2
>  4 0 0  501500  89424     5   0   0   0     0   0   0   0 1561 6544 1818 85 15  0
>  4 0 0  501300  89424  1693   0   0   0  1693   0   1   0 1507 9523 1780 86 14  0
>  3 0 0  501792  89424   876   0   0   0   873   0   3   0 1321 14872 1693 81 19  0
>  1 0 0  501580  89424  2674   0   0   0  2574   0   0   0 1305 10642 1690 88 12  0
>  2 0 0  501580  89424   876   0   0   0   866   0   0   0 1266 12783 2005 58 10 33
>  2 0 0  501580  89424     0   0   0   0     0   0   3   0 1291 1464  996 2  6 92
> 
> 
> but still a slightly higher than on 6.X, but I'm still experiencing
> slowdown of webpage rendering during compilation.
>

The "slightly higher" is my fault, I was running an additional
application compared to the previous test, without this app the result
(the number of cs is identical).

-- 
Marc
Received on Mon Oct 08 2007 - 12:31:18 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:19 UTC