Vulnerabilities / Threats

11/6/2018
05:40 PM
50%
50%

'PortSmash' Brings New Side-Channel Attack to Intel Processors

New vulnerability exposes encryption keys in the first proof-of-concept code.

A new Intel side-channel vulnerability dubbed PortSmash promises to lay encryption keys open to discovery by threat actors.

PortSmash uses characteristics of Simultaneous Multi Threading (SMT), a technique used in Intel processors to run two programs on a single core. In this case, one of the processes sniffs the activity of the other until it's able to work out the timing and details of instructions based on port contentions and their resolution.

The proof of concept, which was created by researchers at Finland’s Tampere University of Technology and the Technical University of Havana, works out the location of data in order to steal an OpenSSL private key from a TLS server.

"Any time you have the opportunity to capture what we assume is secret — in this case a private key — it's serious," says Kevin Bocek, vice president of security strategy at Venafi. In the future when more organizations have complex workloads running on cloud platforms and lack complete control and full visibility, this sort of vulnerability will become even more serious, he notes.

One factor underscoring the seriousness of the vulnerability is the relative ease with which it can be exploited, says Justin Jett, director of audit and compliance at Plixer. Discovering the core on which the TLS server is running is easy with open source tools like taskset or cpuset, he says, and then injecting a process into that same core is quite simple.

"Ultimately, you create malware that will be inserted into the same core [as the server]. This malware then has the capability of getting the decryption key that OpenSSL is generating," he explains.

The issue could become worse, Jett says, "if that site also has deployed some form of single sign-on, then the credentials passed through the Web server could be compromised. And if that single sign-on happens to use Active Directory server, then that could further compromise the systems themselves."

If there's good news here, it's that the same issues currently exist for PortSmash that exist for all similar side-channel vulnerabilities. "This type of attack means that you're already able to execute code on somebody's computer," says Chet Wisniewski, principal research scientist at Sophos. "Typically, once you've already got malicious code running on a computer, why bother executing such a complicated thing in order to potentially steal encryption you probably can already steal?" 

Wisniewski says he believes PortSmash deserves a rating of four on a ten-point scale of severity, but that it could prove more dangerous as time goes on. "A four can turn into a seven or eight down the road. Right now it's too hard for an average criminal to bother with it, but as researchers refine the attack over six-, 12-, 18 months, we've seen all the other attacks suddenly become more important," he explains.

Little Control

There's relatively little that cloud or hosted service customers can do to protect themselves from this sort of vulnerability, researchers say. "Changing the frequency with which TLS keys are replaced is important," Venafi's Bocek says.

"We should certainly be monitoring processes, and making sure that we know there aren't rogue processes that are on the servers that are injecting themselves into specific cores," says Plixer's Jett.

The threat will be problematic for mid-tier organizations without the knowledge or resources to update their applications with newer encryption libraries more resilient to the attack, Wisniewski notes.

"It's inevitable: death, taxes, and more side-channel attacks. I can't protect the chips, but I can protect the TLS certificates," says Bocek. "There will be more side-channel attacks. The sky isn't falling, but it is serious."

Related Content:

 

Black Hat Europe returns to London Dec 3-6 2018  with hands-on technical Trainings, cutting-edge Briefings, Arsenal open-source tool demonstrations, top-tier security solutions and service providers in the Business Hall. Click for information on the conference and to register.

Curtis Franklin Jr. is Senior Editor at Dark Reading. In this role he focuses on product and technology coverage for the publication. In addition he works on audio and video programming for Dark Reading and contributes to activities at Interop ITX, Black Hat, INsecurity, and ... View Full Bio

Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
janjikiu
50%
50%
janjikiu,
User Rank: Apprentice
11/7/2018 | 2:28:36 PM
very nice
very nice
Microsoft President: Governments Must Cooperate on Cybersecurity
Kelly Sheridan, Staff Editor, Dark Reading,  11/8/2018
5 Reasons Why Threat Intelligence Doesn't Work
Jonathan Zhang, CEO/Founder of WhoisXML API and TIP,  11/7/2018
Why Password Management and Security Strategies Fall Short
Steve Zurier, Freelance Writer,  11/7/2018
Register for Dark Reading Newsletters
White Papers
Video
Cartoon
Current Issue
Flash Poll
Online Malware and Threats: A Profile of Today's Security Posture
Online Malware and Threats: A Profile of Today's Security Posture
This report offers insight on how security professionals plan to invest in cybersecurity, and how they are prioritizing their resources. Find out what your peers have planned today!
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2018-6260
PUBLISHED: 2018-11-13
NVIDIA graphics driver contains a vulnerability that may allow access to application data processed on the GPU through a side channel exposed by the GPU performance counters. Local user access is required. This is not a network or remote attack vector.
CVE-2018-16850
PUBLISHED: 2018-11-13
postgresql before versions 11.1, 10.6 is vulnerable to a to SQL injection in pg_upgrade and pg_dump via CREATE TRIGGER ... REFERENCING. Using a purpose-crafted trigger definition, an attacker can cause arbitrary SQL statements to run, with superuser privileges.
CVE-2018-17187
PUBLISHED: 2018-11-13
The Apache Qpid Proton-J transport includes an optional wrapper layer to perform TLS, enabled by use of the 'transport.ssl(...)' methods. Unless a verification mode was explicitly configured, client and server modes previously defaulted as documented to not verifying a peer certificate, with options...
CVE-2018-1792
PUBLISHED: 2018-11-13
IBM WebSphere MQ 8.0.0.0 through 8.0.0.10, 9.0.0.0 through 9.0.0.5, 9.0.1 through 9.0.5, and 9.1.0.0 could allow a local user to inject code that could be executed with root privileges. IBM X-Force ID: 148947.
CVE-2018-1808
PUBLISHED: 2018-11-13
IBM WebSphere Commerce 9.0.0.0 through 9.0.0.6 could allow some server-side code injection due to inadequate input control. IBM X-Force ID: 149828.