linux-kernel - Re: [Xen-devel] [PATCH 4/4] xen/PVH: Remove reserved entry in PVH GDT

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 02 May 2018 09:07:24 -0600
From:   "Jan Beulich" <JBeulich@...e.com>
To:     "Boris Ostrovsky" <boris.ostrovsky@...cle.com>
Cc:     "Roger Pau Monne" <roger.pau@...rix.com>,
        "xen-devel" <xen-devel@...ts.xenproject.org>,
        "Juergen Gross" <jgross@...e.com>, <linux-kernel@...r.kernel.org>
Subject: Re: [Xen-devel] [PATCH 4/4] xen/PVH: Remove reserved entry in
 PVH GDT

>>> On 02.05.18 at 17:06, <boris.ostrovsky@...cle.com> wrote:
> On 05/02/2018 04:26 AM, Jan Beulich wrote:
>>>>> On 01.05.18 at 14:34, <boris.ostrovsky@...cle.com> wrote:
>>> On 05/01/2018 04:00 AM, Roger Pau Monné wrote:
>>>> On Mon, Apr 30, 2018 at 12:23:39PM -0400, Boris Ostrovsky wrote:
>>>>> And without it we can't use _BOOT_XX macros any longer so define new ones.
>>>> Not being that familiar with Linux internals I'm not sure I see the
>>>> benefit of this. Isn't there a risk that some other code is going to
>>>> use the __BOOT_XX defines?
>>> The startup code we are jumping to loads their own GDT and I don't see 
>>> any explicit references to segments.
>> No explicit references to segments isn't enough: You also need to make
>> sure no exceptions at all can occur while loaded selectors and GDT are
>> out of sync - in particular NMI might be of concern here (this isn't PV
>> after all, where not having a callback registered effectively masks NMI).
> 
> How would keeping __BOOT_XX selectors help with NMI? We don't have
> anything set up for NMI handling anyway yet, this is all done in x86
> startup code later.

Oh, you're right - there's no IDT either, so an NMI would yield a triple fault
anyway.

Jan