From mboxrd@z Thu Jan  1 00:00:00 1970
References: <20150824181426.GA21893@hermes.click-hack.org>
From: Jan Kiszka <jan.kiszka@siemens.com>
Message-ID: <55DB6151.8050003@siemens.com>
Date: Mon, 24 Aug 2015 20:24:17 +0200
MIME-Version: 1.0
In-Reply-To: <20150824181426.GA21893@hermes.click-hack.org>
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: 7bit
Subject: Re: [Xenomai] perf nmi on x86
List-Id: Discussions about the Xenomai project <xenomai.xenomai.org>
List-Unsubscribe: <http://xenomai.org/mailman/options/xenomai>,
 <mailto:xenomai-request@xenomai.org?subject=unsubscribe>
List-Archive: <http://xenomai.org/pipermail/xenomai/>
List-Post: <mailto:xenomai@xenomai.org>
List-Help: <mailto:xenomai-request@xenomai.org?subject=help>
List-Subscribe: <http://xenomai.org/mailman/listinfo/xenomai>,
 <mailto:xenomai-request@xenomai.org?subject=subscribe>
To: Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org>, Xenomai <xenomai@xenomai.org>

On 2015-08-24 20:14, Gilles Chanteperdrix wrote:
> Hi Jan,
> 
> running Xenomai on an x86 machine, I noticed that the perf NMI was
> ticking. It seems CONFIG_PERF can not be disabled in the kernel
> configuration.
> 
> So, the first question is: 
> is there a way of disabling the NMI ?
> second question is: 
> if not, what do we do about it? 
> 
> I mean, on the one hand, disabling the NMI is probably mandatory on
> a system where latency matters, but on the other hand, some users
> may want it enabled and do not care for the additional latency.

Does "echo 0 > /proc/sys/kernel/nmi_watchdog" change the picture? It did
so for me on a distro kernel, though re-enabling it again didn't make
the PMIs reappear.

Jan

-- 
Siemens AG, Corporate Technology, CT RTC ITP SES-DE
Corporate Competence Center Embedded Linux