[patch/RFA] multiarch INSTRUCTION_NULLIFIED

[Andrew Cagney]

Randolph Chung wrote:
> Currently only hppa uses this define, but possibly this will be useful
> for other architectures which has conditionally nullified instructions
> (ia64?) as well. This gets us one step closer to full multiarch 
> support.. 
> 
> Tested on hppa-linux with no regressions.

I'm wondering about:

# Return non-zero if the processor is executing a delay slot and a
# further single-step is needed before the instruction finishes.
M::int:single_step_through_delay:struct frame_info *frame:frame

and the commented out code:

   /* If PC is pointing at a nullified instruction, then step beyond
      it so that the user won't be confused when GDB appears to be ready
      to execute it. */

   /*      if (INSTRUCTION_NULLIFIED && currently_stepping (ecs)) */

I suspect that at one stage it was behaving like 
single_step_through_delay, but was then changed to perform the same 
operation unconditionally (i.e., when stopping from a cntrl-c say).  GDB 
trying to helpfully do a few extra steps after a cntrl-c bugs me :-)

Andrew