[<prev] [next>] [day] [month] [year] [list]
Message-Id: <48A575B3.76E4.0078.0@novell.com>
Date: Fri, 15 Aug 2008 11:25:23 +0100
From: "Jan Beulich" <jbeulich@...ell.com>
To: "Ingo Molnar" <mingo@...e.hu>,
"Venkatesh Pallipadi" <venkatesh.pallipadi@...el.com>,
"Thomas Gleixner" <tglx@...utronix.de>, <hpa@...or.com>
Cc: "Jeremy Fitzhardinge" <jeremy@...p.org>,
<linux-kernel@...r.kernel.org>
Subject: Why can't _PAGE_PROTNONE be _PAGE_ACCESSED?
With the overlaying of _PAGE_PSE and _PAGE_PAT there is a latent issue
with the definition of _PAGE_PROTNONE: As soon as _PAGE_PAT gets
actively used, it'll need to become part of _PAGE_CHG_MASK, and hence
_PAGE_PROTNONE cannot use _PAGE_PSE anymore. One option certainly
is to use one of the two remaining unused (available) bits, but it would
seem to me that simply using _PAGE_ACCESSED here ought to work - but
perhaps I'm overlooking some critical aspect?
The background of this is that Xen's dom0 (and similarly DomU-s that have
physical devices assigned) cannot change the setting in PAT_MSR, and
hence has to live with being required to use _PAGE_PAT for establishing
WC mappings. That'll be a requirement for Dom0 and pv pass-through
support in the mainline tree (whenever that gets enabled) anyway, but
I'll have to do it eagerly in our forward-ported Xen tree, which is why I'm
trying to settle early on the route to take here.
Thanks, Jan
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists