You could be excused if you thought Microsoft's month for cybersecurity flaws was July 2021.
There was PrintNightmare in various forms, then HiveNightmare (a completely separate issue that was given the “Nightmare” moniker), and finally PetitPotam (which went down the cute aquatic mammal naming path).
Now, however, it’s Apple’s turn to be in the patch-right-now spotlight, with a somewhat under-announced emergency zero-day fix, just a few days after the company’s last, and much broader, security update.
This one doesn’t have a fancy name, but instead goes simply by CVE-2021-30807, and was reported, according to Apple “by an anonymous researcher”.
Indeed, all we know about it, and all Apple has said so far, is that:
An application may be able to execute arbitrary code with kernel privileges. Apple is aware of a report that this issue may have been actively exploited.
If the crooks knew about it beforehand, it's a zero-day problem, which refers to a fix that was released after the Bad Guys got a head start rather than before they found it out for themselves.
The vulnerability was apparently found in the IOMobileFrameBuffer kernel code, a component that helps userland applications (in other words, unprivileged software) to configure and use your device’s or computer’s display.
The functions supported by IOMobileFrameBuffer help with the management of settings such as video power saving, as well as colour and brightness correction.
We're guessing there's a function in there that could be called in an unexpected and erroneous way, resulting in a buffer overflow or buffer misdirection, in which the kernel failed to check the parameters it was passed, allowing an unprivileged program to shove data into privileged memory where it wasn't supposed to be.
That sort of bug frequently leads to DoS, or denial of service attacks, where a malicious program can deliberately crash the device at will.
Memory corruption bugs sometimes lead to information leakage holes, where a malicious program can read out other people’s data that is supposed to secret.
In the worst cases, however, the ability to make controlled but unauthorised modifications to kernel memory may cause even more serious kernel bugs.
These include elevation of privilege (EoP), where an otherwise uninteresting app suddenly gets the same sort of power as the operating system itself, or even remote code execution (RCE), where an otherwise innocent operation, such as viewing a web page or opening up an image, could trick the kernel into running completely untrusted code that didn’t come from Apple itself.
In particular, when Apple notes that “an application may be able to execute arbitrary code with kernel privileges”, you should assume that an attacker could not only steal your personal data without any visible warnings, but also effectively “jailbreak” your device, thereby bypassing Apple’s protective security boundaries entirely, without so much as a by-your-leave.


0 Comments