Date: Tue, 16 Nov 2004 23:36:08 +0100
From: Ingo Molnar <mingo@elte.hu>
To: Dean Nelson <dcn@sgi.com>
Cc: Chris Wright <chrisw@osdl.org>, akpm@osdl.org,
       linux-kernel@vger.kernel.org
Subject: Re: [Patch] export sched_setscheduler() for kernel module use
Message-ID: <20041116223608.GA27550@elte.hu>
References: <4198F70D.mailxMSZ11J00J@aqua.americas.sgi.com> <20041115105801.T14339@build.pdx.osdl.net> <20041115203343.GA32173@sgi.com> <20041116104821.GA31395@elte.hu> <20041116201841.GA29687@sgi.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20041116201841.GA29687@sgi.com>
User-Agent: Mutt/1.4.1i
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1204
Lines: 26


* Dean Nelson <dcn@sgi.com> wrote:

> > could you make sched_setscheduler() also include a parameter for the
> > nice value, so that ->static_prio could be set at the same time too
> > (which would have relevance if SCHED_OTHER is used)? This would make it
> > a generic kernel-internal API to change all the priority parameters.
> > Looks good otherwise.
> 
> Yeah, I can do that. I'll probably be getting back to you with a
> question or two, if what you're after isn't obvious once I start
> making the changes for the nice parameter.

another potential API would be to use the linear priority range that the
scheduler has internally, from 0 (RT prio 99) to 140 (nice +19). I'm not
sure which solution is the better one. Using the linear priority has the
advantage of not having to pass any policy value - priorities between 0
and 99 implicitly mean SCHED_FIFO, priorities above that would mean
SCHED_NORMAL, a pretty natural and compact interface.

	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/