This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
Home Products Support Community News


Re: [Xen-devel] [PATCH] This patch fixes several issues related to vmxas

To: "Li, Xin B" <xin.b.li@xxxxxxxxx>
Subject: Re: [Xen-devel] [PATCH] This patch fixes several issues related to vmxassist
From: Keir Fraser <Keir.Fraser@xxxxxxxxxxxx>
Date: Sat, 1 Apr 2006 16:57:17 +0100
Cc: Xen Devel <xen-devel@xxxxxxxxxxxxxxxxxxx>
Delivery-date: Sat, 01 Apr 2006 16:01:38 +0000
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
In-reply-to: <59D45D057E9702469E5775CBB56411F10222B3CC@pdsmsx406>
List-help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-id: Xen developer discussion <xen-devel.lists.xensource.com>
List-post: <mailto:xen-devel@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
References: <59D45D057E9702469E5775CBB56411F10222B3CC@pdsmsx406>
Sender: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx

On 1 Apr 2006, at 16:51, Li, Xin B wrote:

Yes, I'll push this one: I like the extra sanity checking. Why are the
checks for zero and greater-than-gdt-limit required (why is it
insufficient to merely check the current mode)?

You got it :-), and I also suspect they are not required.
I just keep it there and seems it's not harmful, when we are sure they
are not needed, we can remove it.

I still think using and maintaining shadow segment state would ultimately be the correct solution. But it sounds like I was mistaken about how big real mode works -- I didn't realise writing a segment register in real mode doesn't update the hidden segment limit register and that's the trick that creates big real mode. That given, the current approach isn't so bad (there's already an acceptable kludge in there to deal with 'stale' hidden base addresses when switching protected-to-real mode) and perhaps we'll get away with leaving the code alone now. It's obviously a pain to revalidate it on a wide range of systems if we start making bigger and more fundamental changes to it. Also, I trust that Leendert knows rather more about the transitory real/protected modes than I do. :-)

 -- Keir

Xen-devel mailing list

<Prev in Thread] Current Thread [Next in Thread>