From mboxrd@z Thu Jan  1 00:00:00 1970
From: Sasha Levin <levinsasha928@gmail.com>
Subject: Coalesced MMIO
Date: Fri, 03 Jun 2011 20:49:44 +0300
Message-ID: <1307123385.13088.18.camel@lappy>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Cc: Avi Kivity <avi@redhat.com>, Pekka Enberg <penberg@kernel.org>,
	Ingo Molnar <mingo@elte.hu>
To: kvm <kvm@vger.kernel.org>
Return-path: <kvm-owner@vger.kernel.org>
Received: from mail-ww0-f42.google.com ([74.125.82.42]:64815 "EHLO
	mail-ww0-f42.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1750969Ab1FCRts (ORCPT <rfc822;kvm@vger.kernel.org>);
	Fri, 3 Jun 2011 13:49:48 -0400
Received: by wwk4 with SMTP id 4so5341038wwk.1
        for <kvm@vger.kernel.org>; Fri, 03 Jun 2011 10:49:47 -0700 (PDT)
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

Hello,

I've tried using KVM_REGISTER_COALESCED_MMIO to register a coalesced
MMIO zone.

ioctl(KVM_CHECK_EXTENSION) for KVM_CAP_COALESCED_MMIO works properly
(and returns 2).

ioctl(KVM_REGISTER_COALESCED_MMIO) with the zone also works fine (and
returns 0).

What I see is that we still get KVM_EXIT_MMIO into that zone, and the
ring never gets used (first==last==0 at all times).

I've went ahead and added some debug output to QEMU, I've noticed that
while qemu registers a coalesced MMIO zone for it's VGA driver,
kvm_flush_coalesced_mmio_buffer() never has any data to flush in the
ring either.

Is it a known issue? Or am I simply not doing something right?

Thanks!

-- 

Sasha.