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 S1752681AbdK0Wl2 (ORCPT <rfc822;w@1wt.eu>);
        Mon, 27 Nov 2017 17:41:28 -0500
Received: from mga03.intel.com ([134.134.136.65]:6670 "EHLO mga03.intel.com"
        rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
        id S1752351AbdK0Wl0 (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
        Mon, 27 Nov 2017 17:41:26 -0500
X-ExtLoop1: 1
X-IronPort-AV: E=Sophos;i="5.44,465,1505804400"; 
   d="scan'208";a="10289462"
Subject: Re: [PATCH 1/5] x86/mm/kaiser: Alternative ESPFIX
To: Peter Zijlstra <peterz@infradead.org>, linux-kernel@vger.kernel.org
References: <20171127223110.479550152@infradead.org>
 <20171127223405.181647306@infradead.org>
Cc: Andy Lutomirski <luto@kernel.org>, Ingo Molnar <mingo@kernel.org>,
        Borislav Petkov <bp@alien8.de>, Brian Gerst <brgerst@gmail.com>,
        Denys Vlasenko <dvlasenk@redhat.com>, "H. Peter Anvin" <hpa@zytor.com>,
        Josh Poimboeuf <jpoimboe@redhat.com>,
        Linus Torvalds <torvalds@linux-foundation.org>,
        Rik van Riel <riel@redhat.com>, daniel.gruss@iaik.tugraz.at,
        hughd@google.com, keescook@google.com, linux-mm@kvack.org,
        michael.schwarz@iaik.tugraz.at, moritz.lipp@iaik.tugraz.at,
        richard.fellner@student.tugraz.at
From: Dave Hansen <dave.hansen@linux.intel.com>
Message-ID: <7ba63732-8824-bb2b-d66b-bddd823de8e8@linux.intel.com>
Date: Mon, 27 Nov 2017 14:41:24 -0800
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101
 Thunderbird/52.4.0
MIME-Version: 1.0
In-Reply-To: <20171127223405.181647306@infradead.org>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
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 11/27/2017 02:31 PM, Peter Zijlstra wrote:
> Change the asm to do the CR3 switcheroo so we can remove the magic
> mappings.
> 
> Since RDI is unused after SWAPGS we can use it as a scratch reg for
> SWITCH_TO_KERNEL. And once we've computed the new RSP (in RAX) we no
> longer need RDI and can again use it as scratch reg for
> SWITCH_TO_USER.

This definitely looks like the right thing.  Either I missed something
obvious before, or Andy's entry rework made this much more obviously
correct to do the simple thing here.