Word leaked out on Tuesday of a new vulnerability in recent versions of Windows that has the potential to unleash the kind of self-replicating attacks that allowed the WannaCry and NotPetya worms to cripple business networks around the world.
This story originally appeared on Ars Technica, a trusted source for technology news, tech policy analysis, reviews, and more. Ars is owned by WIRED’s parent company, Condé Nast.
The vulnerability exists in version 3.1.1 of the Server Message Block 3.1.1 that’s used to share files, printers, and other resources on local networks and over the Internet. Attackers who successfully exploit the flaw can execute code of their choice on both servers and end-user computers that use the vulnerable protocol, Microsoft said in this bare-bones advisory.
The flaw, which is tracked as CVE-2020-0796, affects Windows 10 and Windows Server 2019, which are relatively new releases that Microsoft has invested huge amounts of resources hardening against precisely these types of attacks. Patches aren’t available, and Tuesday’s advisory gave no timeline for one being released. Asked if there was a timeline for releasing a fix, a Microsoft representative said: “Beyond the advisory you linked, nothing else to share from Microsoft at this time.”
In the meantime, Microsoft said vulnerable servers can be protected by disabling compression to block unauthenticated attackers from exploiting the vulnerability against an SMBv3 server. Users can use the following PowerShell command to turn off compression without needing to reboot the machine:
Set-ItemProperty -Path “HKLM:SYSTEMCurrentControlSetServicesLanmanServerParameters” DisableCompression -Type DWORD -Value 1 -Force
That fix won’t protect vulnerable client computers from attack. Microsoft also recommended users block port 445, which is used to send SMB traffic between machines.
Now You See It, Now You Don’t
An advisory published—and then removed—by security firm Fortinet described the vulnerability as “MS.SMB.Server.Compression.Transform.Header.Memory.Corruption.” The pulled advisory said the flaw is the result of a buffer overflow in vulnerable Microsoft SMB servers.
“The vulnerability is due to an error when the vulnerable software handles a maliciously crafted compressed data packet,” Fortinet researchers wrote. “A remote, unauthenticated attacker can exploit this to execute arbitrary code within the context of the application.”
Cisco’s Talos security team also published—and later pulled—its own advisory. It called the vulnerability “wormable,” meaning a single exploit could touch off a chain reaction that allows attacks to spread from vulnerable machine to vulnerable machine without requiring any interaction from admins or users.
“An attacker could exploit this bug by sending a specially crafted packet to the target SMBv3 server, which the victim needs to be connected to,” the removed Talos post said. “Users are encouraged to disable SMBv3 compression and block TCP port 445 on firewalls and client computers. The exploitation of this vulnerability opens systems up to a ‘wormable’ attack, which means it would be easy to move from victim to victim.”
Microsoft’s implementation of SMBv3 introduces a variety of measures designed to make the protocol more secure on Windows computers. The update became more widely used after WannaCry and NotPetya used an exploit developed by—and later stolen from—the National Security agency. Known as EternalBlue, the attack exploited SMBv1 to gain remote code execution and move from machine to machine. Microsoft has similarly hardened Windows 10 and Server 2019 to better withstand exploits, especially those that would otherwise be wormable.
It’s not clear why Microsoft released the sparse details or why both Fortinet and Talos released and then pulled their advisories. The event came on Update Tuesday, which occurs on the second Tuesday of each month, when Microsoft releases a crop of patches to fix various security vulnerabilities.
While CVE-2020-0796 is potentially serious, not everyone said it poses the kind of threat mounted by the SMBv1 flaw that was exploited by WannaCry and NotPetya. Those worms were fueled by the public release of EternalBlue, an exploit that was so reliable it made exploitation a copy-and-paste exercise. Another major contribution to the worms’ success was the near-ubiquity of the SMBv1 at the time. SMBv3, by contrast, is much less used.
SMB is also protected by kernel address space layout randomization, a protection that randomizes the memory locations where attacker code gets loaded in the event a vulnerability is successfully exploited. The protection requires attackers to devise two highly reliable exploits, one that abuses a buffer overflow or other code-execution vulnerability and another that reveals the memory locations of the malicious payload. The protection required Buckeye, an advanced hacker group that exploited the SMBv1 flaw 14 months before the mysterious leak of EternalBlue, to use a separate information disclosure flaw as well.