Hello, I am a Ph.D. student at Purdue University researching kernel compartmentalization. I am currently conducting evaluation for a follow-up paper to one published at NDSS 22 titled 'Preventing Kernel Hacks with HAKC' (see [1]). We are interested in empirically evaluating our compartmentalization policies by determining if targets of exploits (which we refer to as exploit sinks) are placed in a different compartment from code that accesses the target (which we refer to as exploit sources). To that end, we are looking for a set of kernel exploit proofs of concept that we can execute and examine. I realize such a set could be sensitive, and I will follow all safety procedures in the handling and execution of any PoC. In lieu of a set of PoCs, are there any statistics of kernel structures that are targeted by attackers? I would imagine, for example, that struct cred would be heavily targeted, but how often and what other kernel structures are targeted would be invaluable. Thank you for any insight you might have. [1] https://www.ndss-symposium.org/ndss2022/accepted-papers/ -- Derrick McKee Phone: (703) 957-9362 Email: derrick.mckee@xxxxxxxxx
