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 S1751142AbcGMQ3L (ORCPT <rfc822;w@1wt.eu>);
	Wed, 13 Jul 2016 12:29:11 -0400
Received: from mga04.intel.com ([192.55.52.120]:6903 "EHLO mga04.intel.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1750865AbcGMQ3E (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
	Wed, 13 Jul 2016 12:29:04 -0400
X-ExtLoop1: 1
X-IronPort-AV: E=Sophos;i="5.28,358,1464678000"; 
   d="scan'208";a="994521405"
Subject: Re: [PATCH 4/4] x86: use pte_none() to test for empty PTE
To: Julia Lawall <julia.lawall@lip6.fr>, Michal Hocko <mhocko@kernel.org>
References: <20160708001909.FB2443E2@viggo.jf.intel.com>
 <20160708001915.813703D9@viggo.jf.intel.com>
 <20160713151820.GA20693@dhcp22.suse.cz>
 <alpine.DEB.2.10.1607131746570.2959@hadrien>
Cc: linux-kernel@vger.kernel.org, x86@kernel.org, linux-mm@kvack.org,
        torvalds@linux-foundation.org, akpm@linux-foundation.org, bp@alien8.de,
        ak@linux.intel.com, dave.hansen@intel.com,
        David Howells <dhowells@redhat.com>
From: Dave Hansen <dave.hansen@intel.com>
Message-ID: <57866C22.4040402@intel.com>
Date: Wed, 13 Jul 2016 09:28:18 -0700
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101
 Thunderbird/38.8.0
MIME-Version: 1.0
In-Reply-To: <alpine.DEB.2.10.1607131746570.2959@hadrien>
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On 07/13/2016 08:49 AM, Julia Lawall wrote:
> My results are below.  There are a couple of cases in arch/mn10300/mm that
> were not in the original patch.

Yeah, so mn10300 is obviously unaffected by the erratum in question, and
I didn't look for non-x86 architectures for this patch.

But, this code definitely _looks_ like it should be using pte_none(),
especially since mn10300 defines it the same way as x86 (well, as x86
_did_ before this series).

	#define pte_none(x)		(!pte_val(x))

> diff -u -p a/arch/mn10300/mm/cache-inv-icache.c b/arch/mn10300/mm/cache-inv-icache.c
> --- a/arch/mn10300/mm/cache-inv-icache.c
> +++ b/arch/mn10300/mm/cache-inv-icache.c
> @@ -45,11 +45,11 @@ static void flush_icache_page_range(unsi
>  		return;
> 
>  	pud = pud_offset(pgd, start);
> -	if (!pud || !pud_val(*pud))
> +	if (!pud || pud_none(*pud))
>  		return;
> 
>  	pmd = pmd_offset(pud, start);
> -	if (!pmd || !pmd_val(*pmd))
> +	if (!pmd || pmd_none(*pmd))
>  		return;