Accessing the CPU in hooks about unicorn-rs

[isra17]

This look like a guenine issue, either the compiler should pick up the lifetime issue or it should not crash. Are you able to get a simple test case that reliably reproduce the issue?
Thanks!

[isra17]

Got it! Need to Box the callback otherwise Rust might allocate them on the stack...
Test case here: isra17@5cb2c4e

The Unicorn interface should be fixed accordingly.

[isra17]

Disregard last comment, crash come from the Unicorn object pointer being first saved in the callback data. The pointer however is first stored on the init function stack. Returning the object from the function move the value to the parent fram and invalidate the saved pointer. I think this can be fixed by forcing Box<Self> on self in add_*_hook functions.

[oblivia-simplex]

Hey, just catching up with this thread now. Glad to hear I stumbled on a worthwhile bug. I'll pull the fix, and see if I get better mileage out of it. Thanks.