[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20140730102128.GJ19379@twins.programming.kicks-ass.net>
Date: Wed, 30 Jul 2014 12:21:28 +0200
From: Peter Zijlstra <peterz@...radead.org>
To: Morten Rasmussen <morten.rasmussen@....com>
Cc: Yuyang Du <yuyang.du@...el.com>,
"mingo@...hat.com" <mingo@...hat.com>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"pjt@...gle.com" <pjt@...gle.com>,
"bsegall@...gle.com" <bsegall@...gle.com>,
"arjan.van.de.ven@...el.com" <arjan.van.de.ven@...el.com>,
"len.brown@...el.com" <len.brown@...el.com>,
"rafael.j.wysocki@...el.com" <rafael.j.wysocki@...el.com>,
"alan.cox@...el.com" <alan.cox@...el.com>,
"mark.gross@...el.com" <mark.gross@...el.com>,
"fengguang.wu@...el.com" <fengguang.wu@...el.com>
Subject: Re: [PATCH 0/2 v4] sched: Rewrite per entity runnable load average
tracking
On Wed, Jul 30, 2014 at 11:13:31AM +0100, Morten Rasmussen wrote:
> It is a very real problem when user-space uses priorities extensively
> like Android does. Tasks related to audio run at very high priorities
> but only for a very short amount of time, but due the to priority
> scaling their load ends up being several times higher than tasks running
> all the time at normal priority. Hence task load is a very poor
> indicator of utilization.
FWIW Android (which I think does this through binder) is utterly insane
in this regard. But yes, we have to live with it.
Content of type "application/pgp-signature" skipped
Powered by blists - more mailing lists