breakpoint.c patch

Stan Shebs shebs@cygnus.com
Fri Aug 27 17:05:00 GMT 1999


   From: toddpw@wrs.com (Todd Whitesel)
   Date: Fri, 27 Aug 1999 16:46:13 -0700 (PDT)

   I suspect that a lot of the hairier algorithms in wait_for_inferior and
   friends result from the need to disambiguate things when you crunch
   through signals. It's not clear if this can all be pushed down into the
   back ends, or if there needs to be some comparing of notes between the
   generic code and the back ends. The best way to find out is to try it.
   It's got to be better than the current freak of evolution that we have now.

I'd certainly like to see a proposal here.  Pushing things down into
the target has been long-desired, but as your cautionary tale of the
Multi hacker shows, it's not something that you can just do one day.
On the positive side, GDB today has a much better separation of host
and target bits than it had when much of the old execution control
code was written, so things that used to be hard should be somewhat
easier.

I think our recent history shows that we can successfully pull off
some major changes to the underlying structure of GDB, as long as we
spend the time on design and the circulation of design proposals, then
follow that up with the commitment of time to actually make the
change.

								Stan


More information about the Gdb mailing list