The issue here is how the SPARC handles traps/faults/interrupts/etc.
   Rather than pushing a lot of machine state on some kind of kernel
   stack, the way a lot of machines do, the SPARC just does a window save,
   but it ignores the window-invalid mask (%wim) when doing so.  This
   [ ... ]


this is how it works on SPARC <v9.

on v9 there is no %wim (the kernel-level handling of register windows
is entirely different) and the double-fault thing is somewhat lessened
by the multiple trap levels available, but otherwise what der Mouse
wrote in the previous message still basically applies.