From mboxrd@z Thu Jan  1 00:00:00 1970
From: Avi Kivity <avi@redhat.com>
Subject: Re: Running KVM inside a chroot
Date: Tue, 18 Oct 2011 12:29:35 +0200
Message-ID: <4E9D550F.9030400@redhat.com>
References: <CAKYuF5ShsK+kc5ox-NrNso5mLfgWqwA-E5bSeHUEdOHTsWEkcA@mail.gmail.com> <B48FE6AD-C0CB-4606-A31D-57A57AFDF76B@suse.de> <CAKYuF5TG4+5yaVZh9KX0wLOjjg2h01Maz-VOsr2u4BVHzE8i7g@mail.gmail.com> <CAKYuF5THb4AjO6U7bBgUTnzpzNf67eKa41wG9jwWibrcuRu+-w@mail.gmail.com> <4E9B04FD.5030201@redhat.com> <CAKYuF5SFRAZ0feo4M2uJM2EyhKzLYLAocNFRt4j1KyHhwt+GRw@mail.gmail.com> <4E9BFBD2.2000103@redhat.com> <CAKYuF5SxQNk_G0--me4dUXoHW0iSec89P1MzHsrwkepnQCcnrg@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Cc: Alexander Graf <agraf@suse.de>, kvm@vger.kernel.org
To: Jorge Lucangeli Obes <jorgelo@chromium.org>
Return-path: <kvm-owner@vger.kernel.org>
Received: from mx1.redhat.com ([209.132.183.28]:37420 "EHLO mx1.redhat.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1753130Ab1JRK3n (ORCPT <rfc822;kvm@vger.kernel.org>);
	Tue, 18 Oct 2011 06:29:43 -0400
In-Reply-To: <CAKYuF5SxQNk_G0--me4dUXoHW0iSec89P1MzHsrwkepnQCcnrg@mail.gmail.com>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

On 10/17/2011 06:37 PM, Jorge Lucangeli Obes wrote:
> On Mon, Oct 17, 2011 at 2:56 AM, Avi Kivity <avi@redhat.com> wrote:
> > On 10/17/2011 08:10 AM, Jorge Lucangeli Obes wrote:
> >> > What do top/vmstat/kvm_stat say?
> >>
> >> I'm attaching the output of the three commands during a test run
> >> launched form inside the chroot, which was slow as usual. I didn't see
> >> anything too weird on top/vmstat, though I found it odd that once the
> >> VM had booted Chromium OS, QEMU still ate 100% of one core. vmstat
> >> didn't show anything strange, QEMU taking up the memory it's supposed
> >> to. What I don't quite know how to interpret is the output of
> >> kvm_stat, hence the attachment.
> >>
> >>
> >
> > kvm_stat shows insane instruction emulation counts.  Please post a trace
> > as described in http://www.linux-kvm.org/page/Tracing.
>
> I've uploaded a bzip'd trace here:
>
> https://docs.google.com/leaf?id=0B78o7gMWkuFeMzM3MWM5NmUtYTY3My00ZDkxLTljZmUtYjRhMWRhOWVjZTZh&hl=en_US
>
> Almost all the exits show "[FAILED TO PARSE]", would this be a problem?
>

Did you 'make install' trace-cmd?  run it from outside the chroot.

You're showing be a trace of the boot process, please start the trace
after the guest is idle (but consuming lots of cpu).

What host kernel version are you running?

-- 
I have a truly marvellous patch that fixes the bug which this
signature is too narrow to contain.