Red Hat Bugzilla – Bug 157631
CAN-2005-0109 timing attack on OpenSSL with HT
Last modified: 2007-11-30 17:07:18 EST
http://www.daemonology.net/hyperthreading-considered-harmful/ describes a timing
attack against certain processors where a local user can cause an information
leak from applications. This could allow, for example, an untrusted local user
to gain access to portions of cryptographic keys.
Note that for an attack to be sucessful you need to be running on an affected
processor, have a malicious local user, and be doing operations with a valuable
The OpenSSL team are working on a patch that will mitigate this issue and remove
the ability to do timing attacks (of any nature) on private key operations. We
will look at backporting and applying this patch and producing OpenSSL updates.
(Just to clarify, we'll also be taking a look at other packages that deal with
private key operations that may be susceptable to similar attacks)
From a D.J.Bernstein posting on comp.arch:
Osvik and Tromer made clear months ago that hyperthreading needed to be
turned off for security. They've been advertising information disclosure
through hyperthreading since at least February. See, for example, the
``Other People's Cache---HyperAttacks with HyperThreading'' abstract in
Fixed in FC development.
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on the solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.