Subject: Re: [PATCH] sched: avoid large irq-latencies in smp-balancing
From: Eric St-Laurent <ericstl34@sympatico.ca>
To: Steven Rostedt <rostedt@goodmis.org>
Cc: Andrew Morton <akpm@linux-foundation.org>,
       Peter Zijlstra <peterz@infradead.org>, mingo@elte.hu,
       pwil3058@bigpond.net.au, linux-kernel@vger.kernel.org
In-Reply-To: <20071107221007.GA31008@goodmis.org>
References: <1194434956.6289.111.camel@twins>
	 <1194437820.6289.114.camel@twins>
	 <20071107102725.64514fde.akpm@linux-foundation.org>
	 <20071107221007.GA31008@goodmis.org>
Content-Type: text/plain
Date: Wed, 07 Nov 2007 19:24:46 -0500
Message-Id: <1194481486.6606.6.camel@perkele>
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1234
Lines: 35


On Wed, 2007-11-07 at 17:10 -0500, Steven Rostedt wrote:
> > 
> > It would be nice if sched_nr_migrate didn't exist, really.  It's hard to
> > imagine anyone wanting to tweak it, apart from developers.
> 
> I'm not so sure about that. It is a tunable for RT. That is we can tweak
> this value to be smaller if we don't like the latencies it gives us.
> 
> This is one of those things that sacrifices performance for latency.
> The higher the number, the better it can spread tasks around, but it
> also causes large latencies.
> 
> I've just included this patch into 2.6.23.1-rt11 and it brought down an
> unbounded latency to just 42us. (previously we got into the
> milliseconds!).
> 
> Perhaps when this feature matures, we can come to a good defined value
> that would be good for all. But until then, I recommend keeping this a
> tunable.


Why not use the latency-expectation infrastructure?

Iterate under lock until (or before...) the system global latency is
respected.


- Eric

-
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/