From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S932067AbcCJTbp (ORCPT <rfc822;w@1wt.eu>);
	Thu, 10 Mar 2016 14:31:45 -0500
Received: from g9t5009.houston.hp.com ([15.240.92.67]:52367 "EHLO
	g9t5009.houston.hp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1752847AbcCJTbg (ORCPT
	<rfc822;linux-kernel@vger.kernel.org>);
	Thu, 10 Mar 2016 14:31:36 -0500
Message-ID: <1457641451.15454.561.camel@hpe.com>
Subject: Re: runtime regression with "x86/mm/pat: Emulate PAT when it is
 disabled"
From: Toshi Kani <toshi.kani@hpe.com>
To: Borislav Petkov <bp@suse.de>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>,
        Richard Purdie <richard.purdie@linuxfoundation.org>,
        Toshi Kani <toshi.kani@hp.com>,
        Bruce Ashfield <bruce.ashfield@windriver.com>,
        "Hart, Darren" <darren.hart@intel.com>,
        "saul.wold" <saul.wold@intel.com>, linux-kernel@vger.kernel.org
Date: Thu, 10 Mar 2016 13:24:11 -0700
In-Reply-To: <20160310192053.GD2194@pd.tnic>
References: <1457366596.15454.281.camel@hpe.com>
	 <20160307210852.GC26051@windriver.com> <1457393912.15454.419.camel@hpe.com>
	 <20160307235328.GD26051@windriver.com> <1457398578.15454.421.camel@hpe.com>
	 <1457400913.15454.435.camel@hpe.com> <20160310144250.GG23251@windriver.com>
	 <1457628591.15454.542.camel@hpe.com> <20160310172029.GA2194@pd.tnic>
	 <1457640261.15454.551.camel@hpe.com> <20160310192053.GD2194@pd.tnic>
Content-Type: text/plain; charset="UTF-8"
X-Mailer: Evolution 3.18.4 (3.18.4-1.fc23) 
Mime-Version: 1.0
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On Thu, 2016-03-10 at 20:20 +0100, Borislav Petkov wrote:
> On Thu, Mar 10, 2016 at 01:04:21PM -0700, Toshi Kani wrote:
> > I will send a patch that sets PAT disabled when MTRR is disabled.  This
> > will solve the Paul's issue.  His qemu32 model does not support PAT,
> > either.
> 
> It does, see my other mail. We need to figure out first why is
> pat_init() being called as part of MTRR setup.

I am not familiar with PPRO_FEATURES, but shouldn't 'flags' in
/proc/cpuinfo show "pat" when X86_FEATURE_PAT is set?

pat_init() is being called as part of MTRR setup because PAT initialization
requires the same CPU rendezvous operation implemented in the MTRR code.

Thanks,
-Toshi