From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:52536)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <dgilbert@redhat.com>) id 1bubPn-0001Ak-9Y
	for qemu-devel@nongnu.org; Thu, 13 Oct 2016 04:31:00 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <dgilbert@redhat.com>) id 1bubPh-0004hW-Sq
	for qemu-devel@nongnu.org; Thu, 13 Oct 2016 04:30:58 -0400
Received: from mx1.redhat.com ([209.132.183.28]:55132)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <dgilbert@redhat.com>) id 1bubPh-0004h9-MI
	for qemu-devel@nongnu.org; Thu, 13 Oct 2016 04:30:53 -0400
Date: Thu, 13 Oct 2016 09:30:49 +0100
From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
Message-ID: <20161013083048.GA2169@work-vm>
References: <20161012140442.GB15590@stefanha-x1.localdomain>
	<CAJSP0QVrdU7P5CVE4Zu2oF7TK0e9AbBNOrnUDCDvpoG2upmZng@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <CAJSP0QVrdU7P5CVE4Zu2oF7TK0e9AbBNOrnUDCDvpoG2upmZng@mail.gmail.com>
Subject: Re: [Qemu-devel] Async savevm using userfaultfd(2)
List-Id: <qemu-devel.nongnu.org>
List-Unsubscribe: <https://lists.nongnu.org/mailman/options/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.nongnu.org/archive/html/qemu-devel/>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <https://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Stefan Hajnoczi <stefanha@gmail.com>
Cc: qemu-devel <qemu-devel@nongnu.org>, John Snow <jsnow@redhat.com>, "Denis V. Lunev" <den@openvz.org>, Rik van Riel <riel@redhat.com>, Andrea Arcangeli <aarcange@redhat.com>

* Stefan Hajnoczi (stefanha@gmail.com) wrote:
> On Wed, Oct 12, 2016 at 4:04 PM, Stefan Hajnoczi <stefanha@gmail.com> wrote:
> > Perhaps this approach can be prototyped with mprotect and a SIGSEGV
> > handler if anyone wants to get async savevm going.  I don't know if
> > there are any disadvantages to mprotecting guest RAM that the kvm kernel
> > module is using.  Hopefully in-kernel devices and vhost will continue to
> > work.
> 
> I woke up this morning with a strong feeling that a SIGSEGV handler
> won't work with vhost.

YKYBHTLW you wake up with strong feelings about SIGSEGV handlers.

>  The problem is that the QEMU process' SIGSEGV
> handler won't be called when the vhost kernel thread faults.  Now I'm
> wondering whether userfaultfd will work together with vhost.

I think it should, or at least I think all other kernel things end up being
caught by userfaultfd during postcopy.

Dave

> Stefan
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK