Spectre Number 4 Disclosure Raises Fresh Hardware Alarms

If you were waiting for the Spectre vulnerabilities to abate, there's a problem -- another variant of the side channel attack is out there. And the patch Intel has let loose for it will decrease CPU performance by around 8%.

This variant -- CVE-2018-3639 or "Spectre Number 4," as it's commonly called -- depends on Spectre-like speculative execution, as well as speculative execution of memory reads before the addresses of all prior memory writes are known. (Yes, this is très geeky.)

With a technique called side-channel analysis, a local user can then figure out data residing on the machine.

The researchers demonstrated Variant 4 in a language-based runtime environment. While Intel is not aware of a successful browser exploit, the most common use of runtimes, such as JavaScript, is in web browsers.

(Source: iStock)

The variant works on x86 chips, including Intel and AMD chipsets, as well as IBM's Power 8, Power 9 and System Z mainframes. It also appears to affect ARM chips as well, too.

So this is not an Intel-only problem.

Intel has published a description of the vulnerability in its "Q2 2018 Speculative Execution Side Channel Update." The chip maker calls it the Speculative Store Bypass (SSB) variant and assign a Medium severity to it.

But, in a change for Intel, the company is addressing the problem heads-on and what mitigation its solution will bring.

While the beta patch the company has delivered will have no performance hit in the default off configuration, Intel did note that: "If enabled, we've observed a performance impact of approximately 2 to 8 percent based on overall scores for benchmarks like SYSmark® 2014 SE and SPEC integer rate on client1 and server2 test systems."

Gartner analyst Neil MacDonald thinks that the security community should pay attention to this.

"The security community needs to care about these latest announcements for three reasons," McDonald told Security Now, adding:

"First, yet another round of microcode updates will be required to remediate them and these have proven to be the most difficult to roll out. Second, even though the severity is rated medium, we are going to want to address these using a structured, risk-based approach and ultimately apply the OS, VMM and microcode updates. There is no need to panic, but we shouldn't leave these systems unpatched forever unless there is no other option or if the risk of arbitrary code execution is near zero (this is certainly not true on most end user systems). Third, as we expected this is just the latest in what were expect will be a number of new ways to attack the underlying vulnerable design implementation. We are only at the beginning of Spectre/Meltdown class attacks. Expect more patches and firmware updates over the next 12-18 months as researchers and attackers continue to try out new attack methods."

MacDonald thinks we aren't done yet in this cycle, and we may not be. But the security community is going to have to decide when the hardware has been sufficiently hardened to be routinely usable.

Related posts:

— Larry Loeb has written for many of the last century's major "dead tree" computer magazines, having been, among other things, a consulting editor for BYTE magazine and senior editor for the launch of WebWeek.