Received-SPF: pass (google.com: domain of linux-kernel+bounces-99185-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
Date: Mon, 11 Mar 2024 16:07:42 +0000
In-Reply-To: <20240311134202.GQ86322@google.com>
Precedence: bulk
Mime-Version: 1.0
References: <2024030251-CVE-2023-52514-c93d@gregkh> <Ze7d0YCDKdu_LJWc@tiehlicka>
 <20240311134202.GQ86322@google.com>
Message-ID: <Ze8sTnE16lbYuX01@google.com>
Subject: Re: CVE-2023-52514: x86/reboot: VMCLEAR active VMCSes before
 emergency reboot
From: Sean Christopherson <seanjc@google.com>
To: Lee Jones <lee@kernel.org>
Cc: Michal Hocko <mhocko@suse.com>, cve@kernel.org, linux-kernel@vger.kernel.org, 
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>, Juergen Gross <jgross@suse.com>, 
	Andrew Cooper <Andrew.Cooper3@citrix.com>
Content-Type: text/plain; charset="us-ascii"

On Mon, Mar 11, 2024, Lee Jones wrote:
> On Mon, 11 Mar 2024, Michal Hocko wrote:
> 
> > On Sat 02-03-24 22:52:59, Greg KH wrote:
> > > Description
> > > ===========
> > > 
> > > In the Linux kernel, the following vulnerability has been resolved:
> > > 
> > > x86/reboot: VMCLEAR active VMCSes before emergency reboot
> > > 
> > > VMCLEAR active VMCSes before any emergency reboot, not just if the kernel
> > > may kexec into a new kernel after a crash.  Per Intel's SDM, the VMX
> > > architecture doesn't require the CPU to flush the VMCS cache on INIT.  If
> > > an emergency reboot doesn't RESET CPUs, cached VMCSes could theoretically
> > > be kept and only be written back to memory after the new kernel is booted,
> > > i.e. could effectively corrupt memory after reboot.
> > > 
> > > Opportunistically remove the setting of the global pointer to NULL to make
> > > checkpatch happy.
> > > 
> > > The Linux kernel CVE team has assigned CVE-2023-52514 to this issue.
> > 
> > I do not really see the security aspect of this fix. Guests systems
> > shouldn't be able to trigger host reboot nor any untrusted entity should
> > on the host either or this would be a serious security hole.

And not just any reboot either, this only comes into play with something like
`reboot -f`.  Not to mention the impact of the bug is ridiculously theroetical
(I didn't tag the patch for stable@ for a reason).

> > Or am I missing something?
> 
> Thanks for reporting.
> 
> If Sean and/or Paolo agree, we can revoke the CVE for you.

Please do.