Steve Gibson (77:07)
That's right. It's revolting. SOC stands for system on a chip. And the vault in the name stems from a subsystem on a huge number of Dell laptop PCs, more than a hundred different models representing tens of millions of Dell laptops. Because it's the latitude, which is, you know, their. Their headline laptop that the subsystem is called the Control Vault 3. So again, the short news blurb that caused me to look deeper just said a set of vulnerabilities can allow threat actors to take control of tens of millions of Dell laptops. The bugs impact the Control Vault 3 firmware that's used to safely store passwords and biometric data inside a secure chip on Dell Windows laptops. The five bugs, codenamed Revolt, impact more than 100 Dell laptop models. The bugs can be exploited. And this is where this guy is like, whoa. Via a Windows API and don't require or doesn't require elevated privileges, Dell has released firmware updates. Okay, so by far the most worrisome part of the entire statement is the bugs can be exploited via a Windows API and don't require elevated privileges. Most flaws that we encounter in security device firmware are actually kind of obscure. They, you know, they require things like boot time access, or access to the system, motherboards, management interface or something. But here we have a set of flaws that literally any Windows app running on anyone's Dell laptop under their non UAC minimal user privilege account could exploit. So here come the miscreants. As I mentioned, Cisco's Talos Security Group discovered and publicly disclosed this last week. I'm sure their discovery was much earlier, since they waited until Dell had created and tested and published the required firmware updates. The good news is those updates exist. The bad news is that they need to be installed before any of Those more than 100 Dell laptop models, representing tens of millions of physical laptops, will be made safe. So let's see what Cisco's Talos Group disclosed last week. Their report leads with four bullet points. First, Talos reported five vulnerabilities to Broadcom. It's actually a Broadcom chip in the Dell laptop. To Broadcom And Dell affecting both the Control Vault 3 firmware and its associated Windows APIs that we're calling Revault Second Bullet Point 100 plus, meaning more than 100 models of Dell laptops are affected by this vulnerability if left unpatched or you know, until patched. Second or third, the revolt attack can be used as a post compromise persistence technique that can remain even across Windows reinstalls. So it's like a rootkit once it gets in there. I mean worse than a rootkit, it's not even going away after a reinstall. And finally, the Revault attack can also be used as a physical compromise to bypass Windows login and or any local user to gain admin and system privileges. Yikes. They can they continued Dell Control Vault is a hardware based security solution that provides a secure bank that stores as in storage bank that stores your passwords, biometric templates and security codes within the firmware. A daughter board provides this functionality and provides these security features through firmware. Dell refers to the daughter board as a ush, a unified security hub as it's used as a hub to run Control Vault connecting various security peripherals such as the fingerprint reader, smart card reader and NFC reader. So you know that's an elegant design to like have all of those physical devices run through a physically separate standalone board before they get to have any access to the motherboard. Good design, they wrote. The current iterations of the product are called Control Vault 3 and Control Vault 3 Plus and can be found in more than 100 different models of actively supported Dell laptops. And I've got links for all this by the way. At the end of this they said mostly from the business centric Latitude and Precision series, these laptop models are widely used in the cybersecurity industry, in government settings and in challenging environments in their rugged version. Sensitive industries that require heightened security when logging in via for example smart card or NFC are more likely to to find Control Vault devices within their environment as they are necessary to enable that is Control Vault is necessary to enable these enhanced security features. So that's just great, right? It's the machines that are most in need of additional security and would therefore likely most would be more likely to be targets that have had their security dramatically impacted by the discovery of these bugs. That's a big whoops. They said. Today Talos is publishing five CVEs and their associated reports. The vulnerabilities include multiple out of bounds vulnerabilities. In other words, good old buffer flow overflows, an arbitrary free and a Stack overflow, all affecting the CV firmware, you know, the Control Vault firmware. We also reported an unsafe deserialization. You know, that'll be an inter, some, some interpretation of something that affects Control Vaults, Windows API. So there was a bug in the Windows code and four bugs in the control, in the, in the Control Vault firmware. So think about that. To accomplish this, Talos would have had to extract and reverse engineer the proprietary firmware from the Broadcom chip. Dell certainly didn't say, hey, please check the firmware we wrote for our core security chip, which provides all of the most critical physical and biometric security for our most secure laptops. My point is here, it's really a shame that this sort of symbiotic relationship doesn't, that sort of symbiotic relationship doesn't exist between manufacturers and security researchers where manufacturers could be of more help to researchers. How many times have we looked at all the extra and unnecessary effort security researchers have had to go through just to reverse engineer and obtain the same information that the manufacturer already has sitting in a file somewhere. And after all that work, which might, might well come to nothing, right? They might have done, you know, extracted the firmware, reverse engineered it, taken a good hard, long look at it and found no problems. But even after they did, the security researchers say to the manufacturer, hey, we just worked our butts off thanklessly for several months to discover a set of five really horrendous security vulnerabilities that affect tens of millions of your most security essential laptops. You know, it's not good news that they're providing, but it is potentially, you know, heading off a, a, a horrific exploit against all of those laptops. So there's something still very wrong with the way we're doing all of this today. The economics in our system are, are not producing, you know, the right incentives. Cisco's Talos group concluded, writing with a lack of common security mitigations and the combination of some of the vulnerabilities mentioned above. The impact of these findings, that is what they discovered they're trying to create, you know, give us some context here. What this means is significant, they said. Let's highlight two of the most critical attack scenarios we've uncovered. First post, compromise pivot, they said on the Windows side, a non administrative user can interact with the Control Vault firmware using its associated APIs and trigger an arbitrary code execution on the CV firmware. Okay. Now given what we know, it's likely possible for the user, meaning an app, some unwitting user, you runs to load an extra large buffer of executable code into control vaults RAM and then cause that buffer to be executed, thus running their own their own CV firmware code on the control vault in order to get up to some mischief, Talos said. From this vantage point, it becomes possible to leak key material essential to the security of the device, thus gaining the ability to permanently modify its firmware. This creates the risk of a so called implant that could stay unnoticed in the laptop's CV firmware and eventually be used as a pivot back onto the system in the case of a threat actor's post compromise strategy. We show how a tampered CV firmware can be used to hack Windows by leveraging the unsafe deserialization bug mentioned previously and then that that was just one. The second is the physical attack. They said a local attacker with physical access to a user's laptop can pry it open and directly access the USH board over USB with a custom connector so there's a USB access to the chip. From there, all the vulnerabilities described previously become in scope for the attacker without requiring the ability to log into the system or knowing a full disk encryption password. While chassis intrusion can be detected, this is a feature that needs to be enabled beforehand to be effective at warning of a potential tampering and is typically not enabled by default in the bios. Another interesting consequence of this scenario is that if a system is configured to be unlocked with a user's fingerprint, it is get this also possible to tamper with the CV firmware to accept any fingerprint rather than only a legitimate users. So when you think about that, the any fingerprint attack is sort of diabolical. How often does anyone go around asking random people to verify that their fingerprint does not unlock their laptop? Probably not often, perhaps never. The affected user would simply notice that their fingerprint reader had apparently suddenly become much better at accepting their fingerprint than it previously had been. And whereas Cisco says to mitigate these attacks, Talos recommends the following Keep your system up to date to ensure the latest firmware is installed. CV firmware can be automatically deployed via Windows Update, but new firmware usually gets released on the Dell website a few weeks prior, so that's good to know. That suggests that that Windows through Microsoft and through Dell through Microsoft will be pushing this out through Windows Update, so it would be nice to have some way to verify that. I didn't dig into this to see whether that is available, but that's great. If not using any of the security peripherals, the fingerprint reader, the smart card reader and the NFC reader. It is possible to disable the CV services using the service manager and or the CV device using device manager. And that's of course that's standard best security practice, right? If you don't need any of these devices, turn them off, then you don't have to worry about them getting used and hacked behind your back. And finally they said it's it's also worth disabling fingerprint login when risks are heightened, for example leaving one's laptop unattended in a hotel room. Windows also provides enhanced sign in security ess which may help mitigate some of the physical attacks and detect inappropriate CV firmware. Then they said to detect an attack consider the following. Depending upon your laptop model, chassis intrusion detection can be enabled in the computer's bios. This would flag physical tampering and may require enabling a password to clear the alert and restart the computer in the Windows logs. Unexpected crashes of the Windows biometric service or the various credential vault services could be a sign of compromise. And Cisco customers using Cisco Secure Endpoint can be made aware of potential risk with the signature definition loaded by abnormal process. Anyway so Dell's own pages label this critical in all caps and they provide a 36 megabyte not very big download to patch this. Excuse me. It's a Windows executable for any Dell Latitude or Precision laptop owners who would like to be proactive and get this patched. I've got the links in the show notes. Hopefully given that it it was it's previously known Dell has published it. Where are we? Today is the second Tuesday of this month. I don't know which month, which patch Tuesday this may have been pushed out but maybe people are getting it today, maybe they got it last month. I don't know where Cisco timed their their release. So it might be that if you try to install the firmware you'll be told it's already been updated and you don't need to install. But anyway, anybody with a Dell laptop, a Dell Latitude probably would be well served to take a look at this and see about getting this resolved. Make sure that they do have the this latest firmware. Okay so the cus, the the listener of ours, Roscoe who I mentioned briefly last week, who is in the field is responsible for managing and maintaining Microsoft systems wrote he said hi Steve, with respect to the SharePoint on prem patching issue, it's important to understand that the ecosystem can be highly complex and patching can be more difficult than it seems. Office 365 might seem to be an obvious way to resolve all these issues, but it can be problematic too. The Enterprise Office Suite has many components which form an extensive requirements matrix and consists of Windows Server version, Active Directory version, Exchange Server version, Dynamics CRM server version, SharePoint server version Mississippi Project server version, Dynamics Great Plains Accounting and Dynamics Human Resources. He said to name some of the commonly deployed solutions, although some have been withdrawn as on PREM installable components. Again, Leo as you noted, because Microsoft is trying to move everybody up to the cloud, he said the versions of all of these components need to be harmonized in order to have a viable working installation. As a result, in order to update or patch to the latest SharePoint on prem version, the trickle down requirements might extend to updating and patching any or all of the other components in the service stack. In extreme situations, this can result in days or even weeks of applying patches, backing out, applying patches in version order, or do different services first, for example applying a patch to Active Directory first, then to CRM, then finally to SharePoint rather than to SharePoint first. The result can be almost unmaintainable, especially for a small to medium enterprise with limited IT resources. Here's an entirely plausible cascade that demonstrates the deep interconnectedness so the sun is shining, the birds are singing, the grass is green, systems are stable, and everything is beautiful in the world, he writes. Yes, a critical vulnerability is discovered in SharePoint with CVS 9.8 and a patch is available. Well, we know that's exactly what happened, he says. The installed SharePoint version is two patch roll ups behind, so SharePoint has to be brought up to date. The second SharePoint roll up will not run on the currently installed Active Directory, so an Active Directory upgrade is required. The Active Directory upgrade implies a Windows Server upgrade. The new Active Directory version no longer supports the installed Exchange version, which also must be upgraded. The new Active Directory version also deprecates NT Landman authentication, which Microsoft SQL was still using. So Microsoft SQL is also in scope the Exchange and the Exchange end of the Dynamic CRM Exchange API deprecates two methods used for email integration from for mail outs to customers and reception of replies in order for replies to be tagged to the original outgoing message inside CRM. Thus, CRM is now in scope for two roll up installations. The second of these CRM roll ups deprecates an API method being used for integration with a parent company's reporting tools, which are required to report the subsidiary sales pipeline prospects to the relevant stock exchange, which is a legislated requirement the parent company must upgrade their data interchange tooling. Alarm bells. The parent company runs Oracle ebs. Seasoned operators might guess which freight train is heading down the tracks. Installing the roll ups in CRM also breaks three in house customizations which must be redesigned and re implemented. At long last the full cascading set of upgrades has been deployed and the SP roll ups and patch can be installed. He says this is the sort of thing that actually happens which can result in weeks no kidding of disruption to business activities and manual workarounds. Is it any surprise Leo, that everyone just wants to leave everything the way it is?
