Re: CFS Scheduler and real-time tasks

[Török Edvin]

On 5/17/07, Ingo Molnar <mingo@xxxxxxx> wrote:
> note: you ran it on a dual-core system and the SMP load-balancer needs
> time to gain precision in the non-RT test - so i'd suggest to test it
> much longer than 10 seconds: 300 seconds at least.

Thanks, running for 300 seconds shows only a 2% error.
Even running a I/O workload while running the massive_intr test keeps
the same error.
BTW, the system is more 'responsive'. For example starting
applications while running the tests, switching windows, moving
windows is much "smoother" than on default scheduler.

[while true;do cp /usr/portage/distfiles/linux-2.6.21.tar.bz2 . &&
sync && rm linux-2.6.21.tar.bz2 && sync; done]

./massiv_new 9 300
020860  00017782
020858  00017774

RT JOB:020865   00003785

020859  00017048
020862  00017387
020855  00017494
020861  00017245
020863  00017492
020864  00017122


Best regards,
Edwin
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/