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 S932835AbdKALe2 (ORCPT <rfc822;w@1wt.eu>);
        Wed, 1 Nov 2017 07:34:28 -0400
Received: from mail.kernel.org ([198.145.29.99]:38572 "EHLO mail.kernel.org"
        rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
        id S932501AbdKALe1 (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
        Wed, 1 Nov 2017 07:34:27 -0400
DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 8D91C21921
Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=kernel.org
Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=luto@kernel.org
X-Google-Smtp-Source: ABhQp+S+1nP9SiYWyP41CVYYeJ1ZW273rkG8egYi7a/C9kq+CnPFCR7DcSlx9279kv4ZOzUlyIWbK2XUUD0TQpeCzSE=
MIME-Version: 1.0
In-Reply-To: <1acdc314-6568-bada-6bea-4fa6ca21fb91@intel.com>
References: <cover.1509006199.git.luto@kernel.org> <8042a4bd5882599dd8e3f9b17f1f40a2fdda0450.1509006199.git.luto@kernel.org>
 <1acdc314-6568-bada-6bea-4fa6ca21fb91@intel.com>
From: Andy Lutomirski <luto@kernel.org>
Date: Wed, 1 Nov 2017 04:34:05 -0700
X-Gmail-Original-Message-ID: <CALCETrU9qTrwfKW0_f5WzG1QTC2yEFxd40G_z7CtY+wwutiyNQ@mail.gmail.com>
Message-ID: <CALCETrU9qTrwfKW0_f5WzG1QTC2yEFxd40G_z7CtY+wwutiyNQ@mail.gmail.com>
Subject: Re: [PATCH 02/18] x86/asm/64: Split the iret-to-user and
 iret-to-kernel paths
To: Dave Hansen <dave.hansen@intel.com>
Cc: Andy Lutomirski <luto@kernel.org>, X86 ML <x86@kernel.org>,
        Borislav Petkov <bpetkov@suse.de>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        Brian Gerst <brgerst@gmail.com>,
        Linus Torvalds <torvalds@linux-foundation.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On Fri, Oct 27, 2017 at 11:05 AM, Dave Hansen <dave.hansen@intel.com> wrote:
> On 10/26/2017 01:26 AM, Andy Lutomirski wrote:
>> +GLOBAL(restore_regs_and_return_to_usermode)
>> +#ifdef CONFIG_DEBUG_ENTRY
>> +     testl   $3, CS(%rsp)
>> +     jnz     1f
>> +     ud2
>
> A nit from the mere mortals in the audience: Could we start commenting
> or make a constant for the user segment bits in CS?

Yeah.  We have such a define, but it's not currently usable from asm.
Also, we can't do the obvious:

testl $SEGMENT_RPL_MASK, ...
jump_if_not_equal_to_KERNEL_RPL

because that makes no sense in asm :(

>
> Also, it would be nice to explain what's going on here.  Maybe:
>
> /*
>  * We think we are returning to the kernel.  Check the
>  * registers we are about to restore and if we appear to
>  * be returning to userspace, do something that will cause
>  * a fault and hopefully an oops report.
>  */
>
> Otherwise, I really like this change.  It's really hard to figure out
> what the context is in the entry assembly in a lot of cases.  It's a
> place where code reuse actually makes things harder to follow.