Re: fix break, not add future-break

[Tom Tromey <tromey@redhat.com>]

>>>>> "Andrew" == Andrew Cagney <ac131313@redhat.com> writes:

Andrew> I'd like to propose that break be modified so that it behaves
Andrew> something like:
Andrew> 	(gdb) break printf.c:printf
Andrew> 	File "printf.c" not currently known, set breakpoint anyway?

For Insight I think it would be useful to have a version of the
command (or a flag, e.g. "break -future") that never asks.  I imagine
that's true of other GUIs as well, though you'd have to ask those more
familiar with them.

The scenario I'm thinking of is saving breakpoints.  Right now we save
all the breakpoints in the session.  When reloading breakpoints, any
breakpoint that isn't immediately valid is discarded.  This is a major
problem for me (and presumably others), since I do a lot of debugging
of code in shared libraries -- meaning that many of my breakpoints are
lost from session to session.

For breakpoints the user enters by hand, asking seems like a nice
idea.

Tom