linux-kernel - Re: [Xen-devel] Re: [PATCH 06/16] xen: disable PAT

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Sun, 10 May 2009 14:29:28 +0100
From:	Alan Cox <alan@...rguk.ukuu.org.uk>
To:	Keir Fraser <keir.fraser@...citrix.com>
Cc:	"H. Peter Anvin" <hpa@...or.com>,
	Ian Campbell <Ian.Campbell@...citrix.com>,
	Jeremy Fitzhardinge <jeremy@...p.org>,
	the arch/x86 maintainers <x86@...nel.org>,
	Suresh Siddha <suresh.b.siddha@...el.com>,
	"Linux@...or.com" <Linux@...or.com>,
	"Pallipadi, Venkatesh" <venkatesh.pallipadi@...el.com>,
	Kernel Mailing List <linux-kernel@...r.kernel.org>,
	Jeremy Fitzhardinge <Jeremy.Fitzhardinge@...rix.com>,
	Xen-devel <xen-devel@...ts.xensource.com>,
	Jan Beulich <jbeulich@...ell.com>, Ingo Molnar <mingo@...e.hu>,
	Lu <yinghai@...nel.org>, Thomas Gleixner <tglx@...utronix.de>,
	Arjan van de Ven <arjan@...radead.org>
Subject: Re: [Xen-devel] Re: [PATCH 06/16] xen: disable PAT

> > There exists at least one processor erratum where the CPU will use
> > PAT[4-7] when the user requested PAT[0-3].  For those CPUs, it is unsafe
> > for *any* OS to have PAT[4-7] != PAT[0-3].
> 
> Would it be unreasonable for us to disable PAT on such processors?

That would be a regression as far as Linux is concerned. They work
perfectly well providing you stick to four PAT entries (which is enough
for any Linux system). I'm not sure how Windows handles this but that
might also matter for Xen I guess.

> have matching PAT configuration. No elfnote would mean use Xen's existing
> PAT setup (or if that's very dangerous then disable PAT altogether, perhaps
> dependent on CPU model/stepping?).

Hiding it on errata hit processors if the guest cannot support PAT
safely on such processors sounds a good policy and its one being a
hypervisor you can do neatly.

There are quite a few different CPUs with PAT errata. I've no idea why
there are so many errata about that specific bit.

Alan
--
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/