Re: [Lse-tech] Re: [PATCH] O(1) scheduler set_cpus_allowed for non-current tasks

From: Kimio Suganuma (k-suganuma@mvj.biglobe.ne.jp)
Date: Wed Feb 20 2002 - 21:29:43 EST

On Wed, 20 Feb 2002 18:04:49 -0800
Paul Jackson <pj@engr.sgi.com> wrote:

> On Wed, 20 Feb 2002, Kimio Suganuma wrote:
> >
> > CPU hotplug needs to change cpus_allowed in definite time.
> > When a process is sleeping for 100000 seconds, how can we offline
> > a CPU the process belongs?
>
> Good - I figured I'd hear from you on this - thanks.
>
> Are you thinking "definite time" on the order of a second?
> I presume you don't require millisecond response time, and that
> minute response time would be too slow, right?

Exactly.

> And just brainstorming ... if a process is sleeping for a long
> time, and the last cpu it executed on is being taken offline,
> what need is there to wake up the process? Let the process
> stay asleep, and find it a new home when it wakes up for other
> reasons.

In such the case, the waken up process's p->cpu must be changed
by another process or in interrupt, not by itself.
So, we cannot assume that p->cpu, or p->cpus_allowed, must be
changed by itself, right?

> In other words, perhaps the goal of having the smallest,
> simplest, least intrusive, most clearly correct code is more
> important here than waking up a process just to tell it that
> it's last cpu went offline.

Smallest, simplest and correct...
I wish I could figure out such codes. :(

Regards,
Kimi 

-- 
Kimio Suganuma <k-suganuma@mvj.biglobe.ne.jp>

-
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/

This archive was generated by hypermail 2b29 : Sat Feb 23 2002 - 21:00:29 EST