Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [PATCH 1/2] x86/entry/64: Do not clear %rbx under Xen
To:     Andy Lutomirski <luto@kernel.org>
Cc:     LKML <linux-kernel@vger.kernel.org>,
        Dominik Brodowski <linux@dominikbrodowski.net>,
        Ingo Molnar <mingo@redhat.com>,
        "H. Peter Anvin" <hpa@zytor.com>,
        Thomas Gleixner <tglx@linutronix.de>,
        Boris Ostrovsky <boris.ostrovsky@oracle.com>,
        Juergen Gross <jgross@suse.com>,
        xen-devel@lists.xenproject.org, X86 ML <x86@kernel.org>,
        stable <stable@vger.kernel.org>,
        Dan Williams <dan.j.williams@intel.com>,
        Andi Kleen <ak@linux.intel.com>
References: <20180721194909.23903-1-m.v.b@runbox.com>
 <CALCETrX435cC7Hi2aTLVrBecZEEeJ3_TgQnTwbObNrt9MKvFzg@mail.gmail.com>
 <b9ad156c-45e9-da80-9cf0-eabbe5a1bbf2@runbox.com>
 <CALCETrXsNxYP7-w-OoiNdJWspTKCsr0i01e6NiwFk1O5jL6grA@mail.gmail.com>
From:   "M. Vefa Bicakci" <m.v.b@runbox.com>
Message-ID: <ce0cf39e-102e-8a33-6d46-492cd34e9506@runbox.com>
Date:   Sat, 21 Jul 2018 19:37:05 -0400
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101
 Thunderbird/52.9.1
MIME-Version: 1.0
In-Reply-To: <CALCETrXsNxYP7-w-OoiNdJWspTKCsr0i01e6NiwFk1O5jL6grA@mail.gmail.com>
Content-Type: text/plain; charset=windows-1252; format=flowed
Content-Language: en-CA
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 07/21/2018 07:30 PM, Andy Lutomirski wrote:
> On Sat, Jul 21, 2018 at 4:19 PM, M. Vefa Bicakci <m.v.b@runbox.com> wrote:
>> On 07/21/2018 05:45 PM, Andy Lutomirski wrote:
>>>
>>> On Sat, Jul 21, 2018 at 12:49 PM, M. Vefa Bicakci <m.v.b@runbox.com>
>>> wrote:
>>>>
>>>> Commit 3ac6d8c787b8 ("x86/entry/64: Clear registers for
>>>> exceptions/interrupts, to reduce speculation attack surface")
>>>> unintendedly
>>>> broke Xen PV virtual machines by clearing the %rbx register at the end of
>>>> xen_failsafe_callback before the latter jumps to error_exit.
>>>> error_exit expects the %rbx register to be a flag indicating whether
>>>> there should be a return to kernel mode.
>>>>
>>>> This commit makes sure that the %rbx register is not cleared by
>>>> the PUSH_AND_CLEAR_REGS macro, when the macro in question is instantiated
>>>> by xen_failsafe_callback, to avoid the issue.
>>>
>>>
>>> Seems like a genuine problem, but:
>>>
>>>> diff --git a/arch/x86/entry/entry_64.S b/arch/x86/entry/entry_64.S
>>>> index c7449f377a77..96e8ff34129e 100644
>>>> --- a/arch/x86/entry/entry_64.S
>>>> +++ b/arch/x86/entry/entry_64.S
>>>> @@ -1129,7 +1129,7 @@ ENTRY(xen_failsafe_callback)
>>>>           addq    $0x30, %rsp
>>>>           UNWIND_HINT_IRET_REGS
>>>>           pushq   $-1 /* orig_ax = -1 => not a system call */
>>>> -       PUSH_AND_CLEAR_REGS
>>>> +       PUSH_AND_CLEAR_REGS clear_rbx=0
>>>>           ENCODE_FRAME_POINTER
>>>>           jmp     error_exit
>>>
>>>
>>> The old code first set RBX to zero then, if frame pointers are on,
>>> sets it to some special non-zero value, then crosses its fingers and
>>> hopes for the best.  Your patched code just skips the zeroing part, so
>>> RBX either contains the ENCODE_FRAME_POINTER result or is
>>> uninitialized.
>>>
>>> How about actually initializing rbx to something sensible like, say, 1?
>>
>> Hello Andy,
>>
>> Thank you for the review! Apparently, I have not done my homework fully.
>> I will test your suggestion and report back, most likely in a few hours.
>>
>> I have been testing with the next/linux-next tree's master branch
>> (dated 20180720), and I noticed that ENCODE_FRAME_POINTER changes the
>> frame pointer (i.e., RBP) register, as opposed to the RBX register,
>> which the patch aims to avoid changing before jumping to error_exit.
>> It is possible that I am missing something though -- I am not sure about
>> the connection between the RBP and RBX registers.
> 
> Sorry, brain fart on my part.

No problem! :-)

>> The change introduced by commit 3ac6d8c787b8 is in the excerpt below. Would
>> it
>> be valid to state that the original code had the same issue that you
>> referred
>> to (i.e., leaving the RBX register uninitialized)?
> 
> Presumably.
> 
> I would propose a rather different fix:
> 
> https://git.kernel.org/pub/scm/linux/kernel/git/luto/linux.git/commit/?h=x86/pti&id=bb3d76b50c3bc78b67d79cf90d328f38a435c793
> 
> Any chance you could test that and see if it fixes your problem?

Of course; I will report back with the result in a few hours.