>>>>> "Perry" == Perry E Metzger <perry@piermont.com> writes:

    Perry> "Eduardo E. Horvath" writes:
    >> On the sparc port, at least, we could commandeer the clean
    >> window trap to check the stack pointer against some limit and
    >> then panic.  You don't need to worry about multiple traps and
    >> watchdog resets.  (This is not an issue for v9 machines.)
    >> 
    >> Unfortunately this is not a MI solution.

    Perry> That doesn't matter -- I suspect the right thing is MD on
    Perry> most ports.

  Also, a bug found via this way on a port that can check may lead to
solutions applicable to other ports as well.

   :!mcr!:            |  Network security consulting and 
   Michael Richardson |      contract programming
 WWW: mcr@sandelman.ottawa.on.ca. PGP key available.