From mboxrd@z Thu Jan  1 00:00:00 1970
From: Tim Deegan <tim@xen.org>
Subject: Re: HYBRID: max_mapped_pfn in ept_set_entry()
Date: Thu, 29 Mar 2012 10:56:14 +0100
Message-ID: <20120329095614.GA72859@ocelot.phlegethon.org>
References: <20120328170810.332f0b93@mantra.us.oracle.com>
	<1333013354.18810.54.camel@zakaz.uk.xensource.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Return-path: <xen-devel-bounces@lists.xen.org>
Content-Disposition: inline
In-Reply-To: <1333013354.18810.54.camel@zakaz.uk.xensource.com>
List-Unsubscribe: <http://lists.xen.org/cgi-bin/mailman/options/xen-devel>,
	<mailto:xen-devel-request@lists.xen.org?subject=unsubscribe>
List-Post: <mailto:xen-devel@lists.xen.org>
List-Help: <mailto:xen-devel-request@lists.xen.org?subject=help>
List-Subscribe: <http://lists.xen.org/cgi-bin/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xen.org?subject=subscribe>
Sender: xen-devel-bounces@lists.xen.org
Errors-To: xen-devel-bounces@lists.xen.org
To: Ian Campbell <Ian.Campbell@citrix.com>
Cc: "Xen-devel@lists.xensource.com" <Xen-devel@lists.xensource.com>, Keir Fraser <keir.xen@gmail.com>, "Tim Deegan (3P)" <Tim.Deegan@citrix.com>, Stefano Stabellini <Stefano.Stabellini@eu.citrix.com>
List-Id: xen-devel@lists.xenproject.org

At 10:29 +0100 on 29 Mar (1333016954), Ian Campbell wrote:
> On Thu, 2012-03-29 at 01:08 +0100, Mukesh Rathor wrote:
> > I'm trying to figure the right thing to do here. Should I just change
> > the "gfn > p2m->max_mapped_pfn" in ept_get_entry() to check for
> > INVALID_MFN? I really shouldn't be adjusting max_mapped_pfn for MMIO
> > pages, right?
> 
> I'm no expert on the p2m side of things but ept_get_entry says:
>     /* This pfn is higher than the highest the p2m map currently holds */
>     if ( gfn > p2m->max_mapped_pfn )
>         goto out;
> 
> which suggests to me that this is just an optimisation (skipping a
> lookup which can never succeed) and therefore it is appropriate to
> update max_mapped_pfn. After all a 1-1 mapped pfn is still a pfn.

Yes, you should adjust max_mapped_pfn, in ept_set_entry() and in
p2m_set_entry() (p2m-pt.c) so we don't get the same bug reappearing
on AMD.  Instead of checking mfn_valid(), they should check 
(p2mt != p2m_invalid && p2mt != p2m_mmio_dm).

> Other places, like ept_walk_table seem to make a similar optimisation. 
> 
> Unless there is some reason to assume that these functions will never be
> passed an MMIO pfn?

No - it's just that this check predates anything other than RAM and
emulated MMIO.

Tim.