Seems like the high-level lesson is that tagged unions as traditionally implemented can be vulnerable in a manner that pointer/address-focused mitigations don’t affect?
Indeed, a small integer tag seems much more vulnerable than a C++ virtual table pointer.
On modern hardware, how much is the penalty for using magic numbers instead of small indices? (A magic number meaning a constant like 0x85adb9ad instead of 2). The compiler can't optimize switch(it->type) using a jump table, but I suspect that branch prediction and speculation makes this optimization barely relevant.
A jump table isn't likely used for very small unions anyways. The biggest downside that I see is that comparing with zero is slightly cheaper if that is a common value.
At the very end, there is speculation that things like this might get manual inclusion in these mitigations grounded in parts of Ivan Krstić's Black Hat talk from last year.
Assuming this was reposted 'automatically' you should have received an email from HN at the email associated with your HN account telling you that it was going to happen. At least I have received such emails in the past.
The amount of work put into the exploit, the breadth of low-level knowledge involved, and the number of moving parts required for the successful privilege escalation is impressive, to say the least.
I suspect that getting thought this all took months, considering that iOS runs on platforms not exactly open for experimentation of such a kind.
iOS has a lot of layers and mitigations both in software and proprietary hardware that aren't found in other systems. Keep in mind that this story would be 20 % the length on other systems, because "physical memory read/write primitive" would be a total break.
I should note that PPL is not designed to protect against the kinds of attack described in this article; it's really meant to prevent substitution of forged page tables and by coincidence the address chosen by the author ended up being unmappable due to an attempt to protect against virtual memory read/write in the kernel.
Let’s give a experienced group of hackers nearly unlimited budget to find security holes in your competitor’s products.
There’s of course, zero bias in that approach, none whatsoever. /s
If project zero spent even half as much time finding and fixing security exploits in Android as they do just finding exploits in IOS, Google would have a truly competitive product from a security perspective.
> News and updates from the Project Zero team at Google
>For the last several years, nearly all iOS kernel exploits ...
Damn, Googlers... Still incapable of explaining things to people. What the hell is Project Zero? Why don't you explain it in your tagline? If not, why not at the beginning of the article?
Edit: this sentence in the "about" section does the job:
> Project Zero is a team of security researchers at Google who study zero-day vulnerabilities in the hardware and software systems that are depended upon by users around the world
I suggest you change the tagline to something like "Studying zero-day vulnerabilities at Google".
