Content-Type: multipart/mixed; boundary="===============8568375853990138157==" MIME-Version: 1.0 From: Fengguang Wu To: lkp@lists.01.org Subject: Re: increased vmap_area_lock contentions on "n_tty: Move buffers into n_tty_data" Date: Tue, 17 Sep 2013 10:56:42 +0800 Message-ID: <20130917025642.GA29664@localhost> In-Reply-To: <5237C183.7040308@hurleysoftware.com> List-Id: --===============8568375853990138157== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable On Mon, Sep 16, 2013 at 10:42:11PM -0400, Peter Hurley wrote: > On 09/12/2013 11:38 PM, Fengguang Wu wrote: > >On Thu, Sep 12, 2013 at 08:17:00PM -0700, Greg KH wrote: > >>On Fri, Sep 13, 2013 at 08:51:33AM +0800, Fengguang Wu wrote: > >>>Hi Peter, > >>> > >>>FYI, we noticed much increased vmap_area_lock contentions since this > >>>commit: > >> > >>What does that mean? What is happening, are we allocating/removing more > >>memory now? > > > >// leave this question to Peter and Tejun. :) > > > >>What type of load were you running that showed this problem? > > > >The increased contentions and lock hold/wait time showed up in a > >number of test cases. > = > How is the raw data acquired? > = > For example, is the test apparatus reading from /proc/vmallocinfo > (which claims the vmap_area_lock) during the test execution? We didn't read /proc/vmallocinfo and read /proc/lock_stat once when the test ends. > I'm wondering if a Heisenberg effect is happening here. Thanks, Fengguang --===============8568375853990138157==--