Re: [PATCH] 2.4-ac: sparc64 support for O(1) scheduler

Ingo Molnar (mingo@elte.hu)
Sun, 16 Jun 2002 17:19:34 +0200 (CEST)


On 14 Jun 2002, Robert Love wrote:

> > Ummm what is with all of those switch_mm() hacks? Is this an attempt
> > to work around the locking problems? Please don't do that as it is
> > going to kill performance and having ifdef sparc64 sched.c changes is
> > ugly to say the least.
> >
> > Ingo posted the correct fix to the locking problem with the patch
> > he posted the other day, that is what should go into the -ac patches.
>
> I am explicitly refraining from sending Alan any code that is not
> well-tested in 2.5 and my machines first. As Ingo's new switch_mm()
> bits are not even in 2.5 yet, [...]

Linus applied them already, they will be in 2.5.22. They fix real bugs and
i've seen no problems on my testboxes. Those bits are a must for SMP x86
and Sparc64 as well, there is absolutely no reason to selectively delay
their backmerge. Besides the last task_rq_lock() optimization which got
undone in 2.5 already, all the recent scheduler bits i posted are needed.

Ingo

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