[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20170710144609.GD31832@tassilo.jf.intel.com>
Date: Mon, 10 Jul 2017 07:46:09 -0700
From: Andi Kleen <ak@...ux.intel.com>
To: Peter Zijlstra <peterz@...radead.org>
Cc: Aubrey Li <aubrey.li@...el.com>, tglx@...utronix.de,
len.brown@...el.com, rjw@...ysocki.net, tim.c.chen@...ux.intel.com,
arjan@...ux.intel.com, paulmck@...ux.vnet.ibm.com,
yang.zhang.wz@...il.com, x86@...nel.org,
linux-kernel@...r.kernel.org, Aubrey Li <aubrey.li@...ux.intel.com>
Subject: Re: [RFC PATCH v1 00/11] Create fast idle path for short idle periods
On Mon, Jul 10, 2017 at 10:46:47AM +0200, Peter Zijlstra wrote:
> On Mon, Jul 10, 2017 at 09:38:30AM +0800, Aubrey Li wrote:
> > We measured 3%~5% improvemnt in disk IO workload, and 8%~20% improvement in
> > network workload.
>
> Argh, what a mess :/
The mess is really the current idle entry path. People keep putting more
and more slow stuff there, not realizing it's really a critical fast path.
>
> So how much of the gain is simply due to skipping NOHZ? Mike used to
> carry a patch that would throttle NOHZ. And that is a _far_ smaller and
> simpler patch to do.
Have you ever looked at a ftrace or PT trace of the idle entry?
There's just too much stuff going on there. NOHZ is just the tip
of the iceberg.
-Andi
Powered by blists - more mailing lists