exploitation_cleanup breaks kernel r/w despite stating it shouldn't

[opa334]

I call exploit_get_krw_and_kernel_base, then try to do a few read and write operations.

After some time, r/w primitives break and kread64 starts to return 0.

If I comment out the call to exploitation_cleanup inside exploit_get_krw_and_kernel_base and instead call it manually after being done using the primitives, everything works.

iPhone 13 Pro, 15.1.1

[potmdehex]

Have to keep nation state actors on their toes a bit.

FilzaEscaped 15 is an example of a project based on multicast_bytecopy (afaik there's no single credit/mention, but you'll see if you open FilzaEscaped.dylib that there isn't much effort to hide the fact that it is multicast_bytecopy - the author hasn't contacted or as much as followed me so to be clear I'm not involved in any way). They moved the cleanup to the end, which you could do for most purposes. This project may or may not be updated in the future.

[opa334]

Yeah, I only figured this out by reversing FilzaEscaped.

No worries, just figured I'd note this here because it lost me a lot of time.