Add hardware stack overflow checking to core and sys

It is possible to set a hardware breakpoint that will cause a fault and stack dump when the stack grows beyond its limits. This is preferable to the current behavior, which is "crash randomly some time later" should the stack grow into the heap (or into the system stack).

A couple gists should help get it started:
https://gist.github.com/earlephilhower/f9f8fa2420ca03b82395da142ad5fc08
https://gist.github.com/earlephilhower/6c00bccb2dfa4c09ef05b1f091e6e763