Scheduler framework patch
Julian Elischer
julian at elischer.org
Fri Aug 27 12:03:38 PDT 2004
Florian C. Smeets wrote:
> Julian Elischer wrote:
>
>> I have made a patch that changes the scheduelr framework to further
>> abstract teh scheduler.
>>
>> it is at http://www.freebsd.org/~julian/f.diff
>> and should apply agains -current (and probably, though I have not
>> tested it yet, against 5.3)
>>
>> there should be no real change in teh way teh system appears to operate.
>> If you are feeling bored you might try applying it to a kernel tree
>> and testing it.
>
>
> Hi Julian,
>
> seems to work, but there is something wrong with the load calculation:
hmm I'll look at that tonight..
>
>
> last pid: 4227; load averages: 576.70, 884.21, 460.30 up
> 0+00:15:10 12:05:41
> 141 processes: 3 running, 112 sleeping, 1 zombie, 25 waiting
> CPU states: 0.8% user, 0.0% nice, 6.9% system, 1.5% interrupt,
> 90.8% idle
> Mem: 137M Active, 38M Inact, 50M Wired, 9576K Cache, 34M Buf, 864K Free
> Swap: 256M Total, 8512K Used, 248M Free, 3% Inuse
>
> PID USERNAME PRI NICE SIZE RES STATE TIME WCPU CPU COMMAND
> 11 root 171 0 0K 12K RUN 12:37 89.06% 89.06% idle
> 1458 flo 20 0 62816K 49912K kserel 0:22 0.00% 0.00%
> thunderbird-bin
> 819 flo 76 0 29592K 16904K select 0:12 0.00% 0.00% Xorg
> 1777 flo 20 0 93240K 44076K kserel 0:08 0.00% 0.00%
> soffice.bin
> 1069 flo 20 0 40436K 23956K kserel 0:07 0.00% 0.00%
> firefox-bin
>
> The fan of my laptop is goning crazy with this patch ;-) This was
> tested with -CURRENT sources as of today + your patch. I'm using ULE
> with libpthread.
I'm not sure how the fan can be influenced by this, but I'll keep it in
mind when looking into it..
how is performance?
thanks for trying it..
>
>
> Regards,
> flo
> _______________________________________________
> freebsd-current at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-current
> To unsubscribe, send any mail to
> "freebsd-current-unsubscribe at freebsd.org"
More information about the freebsd-current
mailing list