RE: [RESEND RFC PATCH] x86/bugs: Add "unknown" reporting for MMIO Stale Data

[David Laight]

From: Pawan Gupta
> Sent: 29 July 2022 03:29
> 
> On Thu, Jul 28, 2022 at 02:00:13PM +0200, Borislav Petkov wrote:
> > On Thu, Jul 14, 2022 at 06:30:18PM -0700, Pawan Gupta wrote:
> > > Older CPUs beyond its Servicing period are not listed in the affected
> > > processor list for MMIO Stale Data vulnerabilities. These CPUs currently
> > > report "Not affected" in sysfs, which may not be correct.

I looked this up....

The mitigations seem to rely on unprivileged code not being able
to do MMIO accesses.
That isn't true, device drivers can mmap PCIe addresses directly
into user program address space.
While unlikely, there is no reason this can't be supported for
non-root processes.
So if the underlying hardware doesn't correctly validate the
byte enables then stale data can be read.

It has to be said that I can't actually imagine getting anything
useful unless you have co-operating processes using it as a
security bypass side channel.

	David

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)