Re: m68k stub and DECR_PC_AFTER_BREAK

[Jim Blandy <jimb at cygnus dot com>]

> Anything, anything.  Just as long as DECR_PC_AFTER_BREAK is eleminated
> from the high level code :-)

Just so you know: Kevin Buettner and I had a conversation about
eliminating DECR_PC_AFTER_BREAK a while ago.  I think Kevin said he'd
look at the problem as time arose.

The problem is that it's not easy to shift a responsibility presently
handled by the core down to the individual architectures.  You need to
go back and fix each relevant tdep file.  Furthermore, there's no
standard way in GDB for an architecture to intercept and adjust
information as it is received from the target.

Fortunately, there are only a few architectures which actually define
DECR_PC_AFTER_BREAK, and Cygnus has an instance of each lying around,
so we have the ability to do the necessary testing.