From mboxrd@z Thu Jan  1 00:00:00 1970
From: Arnd Bergmann <arnd@arndb.de>
Subject: Re: [RFC PATCH 0/3] generic hypercall support
Date: Thu, 7 May 2009 22:00:21 +0200
Message-ID: <200905072200.22007.arnd@arndb.de>
References: <20090505132005.19891.78436.stgit@dev.haskins.net> <4A032472.4030106@redhat.com> <4A03259B.3050500@gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-15
Content-Transfer-Encoding: QUOTED-PRINTABLE
Cc: Avi Kivity <avi@redhat.com>, Gregory Haskins <ghaskins@novell.com>,
	Chris Wright <chrisw@sous-sol.org>,
	linux-kernel@vger.kernel.org, kvm@vger.kernel.org,
	Anthony Liguori <anthony@codemonkey.ws>
To: Gregory Haskins <gregory.haskins@gmail.com>
Return-path: <kvm-owner@vger.kernel.org>
Received: from moutng.kundenserver.de ([212.227.126.186]:62901 "EHLO
	moutng.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1753298AbZEGUBs convert rfc822-to-8bit (ORCPT
	<rfc822;kvm@vger.kernel.org>); Thu, 7 May 2009 16:01:48 -0400
In-Reply-To: <4A03259B.3050500@gmail.com>
Content-Disposition: inline
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

On Thursday 07 May 2009, Gregory Haskins wrote:
> I guess technically mmio can just be a simple access of the page whic=
h
> would be problematic to trap locally without a PF. =A0However it seem=
s
> that most mmio always passes through a ioread()/iowrite() call so thi=
s
> is perhaps the hook point. =A0If we set the stake in the ground that =
mmios
> that go through some other mechanism like PFs can just hit the "slow
> path" are an acceptable casualty, I think we can make that work.
>=20
> Thoughts?

An mmio that goes through a PF is a bug, it's certainly broken on
a number of platforms, so performance should not be an issue there.

Note that are four commonly used interface classes for PIO/MMIO:

1. readl/writel: little-endian MMIO
2. inl/outl: little-endian PIO
3. ioread32/iowrite32: converged little-endian PIO/MMIO
4. __raw_readl/__raw_writel: native-endian MMIO without checks

You don't need to worry about the __raw_* stuff, as this should never
be used in device drivers.

As a simplification, you could mandate that all drivers that want to
use this get converted to the ioread/iowrite class of interfaces and
leave the others slow.

	Arnd <><