Re: Results of schedule() switching test with my patches ...

Davide Libenzi (davidel@maticad.it)
Thu, 10 Jun 1999 15:16:24 +0200

>
>On Wed, 9 Jun 1999, Davide Libenzi wrote:
>
>> The test program is very simple, it creates 450 threads that do nothing
>> but increment a counter and call syscall(SYS_sched_yield).
>
>> My results are:
>
>> Average gain = + 56 % !!!!
>
>what is the speed difference between the original and the patched kernel
>if you have only 2 threads?
>
>-- mingo
>

Hi mingo,

what kind of question is it ?
Have You seen the patch ?
It is clear that if You have 2 threads the patch give results near to the
original.
But if You have a lot of task ( hi loaded servers ) the results are _good_.
My patch avoid schedule() to scan 450 tasks ( more of them with 0 goodness )
to find the one good to run !!
More, one a great number of cases it is sufficent one test to get the good
one.

Cheers,
Davide.

--
"Debian, the Freedom in Freedom."








-
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/